UNITED STATES
SECURITIES AND EXCHANGE COMMISSION

FORM 10-K

Annual Report Pursuant to Section 13 or 15(d)
of the Securities Exchange Act of 1934
For the Fiscal Year Ended December 31, 2004

Commission File Number 000-50642

MEMORY PHARMACEUTICALS CORP.

100 Philips Parkway
Montvale, New Jersey 07645
(201) 802-7100
(Address, including zip code, and telephone number,
including area code, of registrant’s principal executive offices)

Securities registered pursuant to Section 12(b) of the Act:
None

Securities registered pursuant to Section 12(g) of the Act:
Common Stock $0.001 par value

     Indicate by check mark whether the registrant (1) has filed all reports required to be filed by Section 13 or 15(d) of the Securities Exchange Act of 1934 during the preceding 12 months (or for such shorter period that the registrant was required to file such reports), and (2) has been subject to such filing requirements for the past 90 days.

þ  Yes     o  No

     Indicate by check mark if disclosure of delinquent filers pursuant to Item 405 of Regulation S-K is not contained herein, and will not be contained, to the best of registrant’s knowledge, in definitive proxy or information statements incorporated by reference in Part III of this Form 10-K or any amendment to this Form  10-K.  þ

     Indicate by checkmark whether the registrant is an accelerated filer (as defined in Exchange Act Rule 12b-2):

     The aggregate market value of the voting stock held by non-affiliates of the registrant as of June 30, 2004 was approximately $137,988,587 (based on the last reported sale price on the NASDAQ National Market on that date). As of March 28, 2005 the registrant had 20,746,923 shares of common stock, par value $0.001 per share, outstanding. The registrant does not have any non-voting stock outstanding.

DOCUMENTS INCORPORATED BY REFERENCE

     Portions of the Registrant’s Proxy Statement for the 2005 Annual Meeting of Stockholders, which is to be filed subsequent to the date hereof, are incorporated by reference into Part III of this Annual Report on Form 10-K.

MEMORY PHARMACEUTICALS CORP.
2004 FORM 10-K ANNUAL REPORT

TABLE OF CONTENTS

This Annual Report on Form 10-K, including the sections labeled Management’s Discussion and Analysis of Financial Condition and Results of Operations, contains forward-looking statements that you should read in conjunction with the financial statements and notes to financial statements that we have included elsewhere in this report. These statements are based on our current expectations, assumptions, estimates and projections about our business and our industry, and involve known and unknown risks, uncertainties, and other factors that may cause our or our industry’s results, levels of activity, performance or achievements to be materially different from any future results, levels of activity, performance or achievements expressed or implied in, or contemplated by, the forward-looking statements. We generally identify these statements by words or phases such as “believe,” “anticipate,” “expect,” “intend,” “plan,” “will,” “may,” “should,” “estimate,” “predict,” “potential,” “continue,” or the negative of such terms or other similar expressions. Our actual results and the timing of events may differ significantly from the results discussed in the forward-looking statements, and you should not place undue reliance on these statements. Factors that might cause such a difference include those discussed below under the heading “Risk Factors,” as well as those discussed elsewhere in this Annual Report on Form 10-K. We disclaim any intent or obligation to update any forward-looking statements as a result of developments occurring after the period covered by this report or otherwise.

PART I

Item 1. Business.

OVERVIEW

We are a biopharmaceutical company focused on the discovery and development of innovative drug candidates for the treatment of a broad range of central nervous system (CNS), conditions that exhibit significant impairment of memory and other cognitive functions. These conditions include neurological diseases associated with aging, such as Alzheimer’s disease, vascular dementia and mild cognitive impairment (MCI), and also include certain psychiatric disorders such as schizophrenia and depression. It is well recognized that memory loss and other cognitive impairments are key characteristics of Alzheimer’s disease and other dementias. It is less widely recognized, however, that memory loss and other cognitive impairments frequently occur in psychiatric disorders such as schizophrenia and depression.

Although therapies for the treatment of Alzheimer’s disease, schizophrenia and depression have been available for a number of years, many of the approved drugs for these disorders are not effective in a large number of patients and can produce significant side effects. Our potential CNS therapies are designed to address biological targets within the cellular pathways that are currently thought to be critically involved in memory formation and other cognitive functions. We believe that this approach could be effective in treating many major neurological and psychiatric disorders. We also believe that the relative specificity of these biological targets provides an opportunity to develop drugs with reduced side effect profiles.

Through pioneering research conducted over more than 30 years, Nobel Laureate Dr. Eric Kandel, one of our scientific founders, identified critical cellular pathways and biological targets involved in memory formation. This research, which was published in the 1990s, served as the cornerstone of our scientific foundation.

In order to identify and optimize promising drug candidates quickly and efficiently we combine:

	Ø	our extensive knowledge of the pathways we believe are involved in memory formation and other cognitive functions;
	Ø	our understanding of neurological and psychiatric disorders;
	Ø	an interdisciplinary drug discovery and development approach; and
	Ø	our focus on conducting in vivo and safety screening at early stages of the drug discovery process.

Two of our drug candidates are currently in clinical development for Alzheimer’s disease and one of our drug candidates is in clinical development for schizophrenia. We also have drug candidates undergoing preclinical testing for Alzheimer’s disease and depression. Our drug discovery pipeline includes compounds that we are optimizing for the treatment of these and other diseases.

1

We seek to leverage our pipeline of early development candidates through collaborations with leading pharmaceutical and biotechnology companies. To date, we have secured two separate collaborations with F. Hoffman – La Roche Ltd. / Hoffman – La Roche, Inc., or Roche, one for the development of our PDE4 inhibitors, including MEM 1414 and MEM 1917, and the other for the development of our nicotinic alpha-7 partial agonists, including MEM 3454 and MEM 63908. As of December 31, 2004, we had received $45.4 million in upfront license fees, payments for research and development services, milestone payments and an equity investment under these collaborations.

MARKET OVERVIEW

According to the World Health Organization (WHO), over 180 million people worldwide suffer from CNS disorders that exhibit significant impairment of memory and other cognitive functions. These disorders include neurological diseases associated with aging, such as Alzheimer’s disease, vascular dementia and MCI, and certain psychiatric diseases including schizophrenia and depression. The cognitive deficits associated with these disorders result in symptoms ranging from mild impairment of short-term and long-term memory to the inability to engage in cogent conversation and perform routine tasks. We expect the market for drugs treating these diseases to grow significantly over the next several decades as the baby boomer generation ages, life expectancies increase and improved drugs and diagnostic techniques are developed to address these diseases.

Alzheimer’s disease is a degenerative neurological disorder that progressively impairs a person’s cognitive function and gradually destroys the brain. It is the number one cause of dementia, and there is no known cure for Alzheimer’s disease. The period of time from onset until patient death averages eight years and can be as long as 20 years. The Alzheimer’s Association reports that approximately 4.5 million people in the US suffer from Alzheimer’s disease, which represents approximately 10% of the US population over the age of 65 and nearly half of those over the age of 85. According to the Alzheimer’s Association, by 2050, the number of Americans with Alzheimer’s disease could range from approximately 11 million to 16 million. National direct and indirect costs of caring for individuals with Alzheimer’s disease are at least $100 billion a year, according to estimates used by the Alzheimer’s Association and the National Institute on Aging.

Schizophrenia is a chronic psychiatric disorder characterized by difficulties differentiating between real and imaginary experiences, thinking logically and managing emotional responses to everyday social situations. According to IMS Health Incorporated, 2004 antipsychotics sales were $14.1 billion worldwide. There are currently no drugs approved for the treatment of the cognitive dysfunction associated with schizophrenia.

Depression is a psychiatric condition that is characterized by excessive sadness, lack of energy and loss of interest in day-to-day activities. In addition, cognitive deficits, such as difficulty concentrating, remembering and making decisions, are common complaints of those suffering from the disease and are now recognized as elements of depression. Depression can last for many years and can require long periods of treatment. According to IMS Health, 2004 antidepressant sales were $20.3 billion worldwide. There are currently no drugs approved for the treatment of the cognitive dysfunction associated with depression.

Vascular dementia is the second leading cause of dementia and often displays symptoms similar to those of Alzheimer’s disease. It is caused by one or more strokes, the slow clogging of the arteries in the brain, traumatic brain injury or other disruptions of blood flow to the brain. There are currently no drugs approved for the treatment of vascular dementia.

MCI is a neurological disorder characterized by greater memory impairment than normally expected with aging but without impaired judgment or reasoning or other symptoms of dementia. Clinical researchers have reported that an estimated 80% of MCI patients progress to Alzheimer’s disease within six years of diagnosis. In 2001, the FDA acknowledged MCI as a distinct clinical indication, which allows for the conduct of trials of compounds for the MCI patient population. There are currently no drugs approved for the treatment of MCI.

2

SCIENTIFIC BACKGROUND

Central nervous system and memory

The CNS is comprised of a network of nerve cells, known as neurons, that enable sensation, memory, emotion and other cognitive functions. Neurons are highly specialized cells that are capable of communicating with each other through biochemical transmission across junctions called synapses. For this transmission to occur, neurons secrete chemicals, known as neurotransmitters, that bind to receptors on a neighboring neuron. Serotonin, dopamine and acetylcholine (ACh) are examples of neurotransmitters. Neurotransmitter signaling can lead to the activation of a specific class of molecules known as second messengers, which can both relay electrical signals and amplify their strength. Cyclic adenosine monophosphate, or cAMP, cyclic guanosine monophosphate, or cGMP and calcium are examples of second messengers. Second messengers are known to play a key role in many intracellular processes of direct relevance to the formation and stabilization of memories.

Coordinated communication across synapses is essential for the formation of both short-term memories, which last for minutes or hours, and long-term memories, which last for days and years. The formation of short-term memories appears to result from the transient release of neurotransmitters across existing synaptic connections. The formation of long-term memories, however, not only requires synaptic neurotransmission, but also a subsequent cascade of cellular events that culminates in protein synthesis, activation of existing synaptic connections and the formation of new synaptic connections. This cascade of events takes place along pathways within the neuron.

Through pioneering research conducted over more than 30 years, Nobel Laureate Dr. Eric Kandel, a founder of the Company, identified critical cellular pathways involved in the formation of both short-term and long-term memories. In addition, Dr. Kandel identified specific targets that are believed to be critical to the process of long-term memory formation and certain classes of compounds that can have restorative effects on the impairment of long-term memory. These findings were published in the 1990s and served as the cornerstone of our scientific foundation.

Memory and cognitive disorders

It is well recognized that memory loss and other cognitive impairments are key characteristics of Alzheimer’s disease and other dementias. It is less widely recognized, however, that memory loss and other cognitive impairments frequently occur in psychiatric disorders such as schizophrenia and depression. For example, many people who have Alzheimer’s disease tend to suffer from depression and, conversely, depressed patients tend to experience significant cognitive decline. Symptoms such as difficulty concentrating, remembering and making decisions are recognized as elements of depression and Alzheimer’s disease. Likewise, common manifestations of schizophrenia and Alzheimer’s disease include the inability to correctly process new information, retrieve information and react appropriately to environmental stimuli. The hallucinations and delusions of schizophrenic people can be traced to misprocessed information or the inability to filter out and decipher common background noises. In recognition of the possible involvement of cognitive dysfunction in schizophrenia, the National Institute of Mental Health, or NIMH, has established an initiative to address the impairments in cognition that appear to be present in many schizophrenic patients. This initiative is based in part on clinical evidence that remediation of these cognitive symptoms is a major determinant and predictor of the success of long-term treatment of schizophrenic patients.

UNMET THERAPEUTIC NEEDS

Therapies for the treatment of Alzheimer’s disease, schizophrenia and depression have been available for a number of years. However, many of the approved drugs for these diseases are not effective in a large number of patients, can produce significant side effects and, in the case of Alzheimer’s disease, lose their effectiveness over time. For example, only about 50% of people taking acetylcholinesterase inhibitors, the most commonly prescribed drug class for Alzheimer’s disease, show even modest improvements in the major symptoms of the disease. Moreover, the benefits derived from taking these drugs typically last for only six to nine months. Of patients taking selective serotonin reuptake inhibitors, or SSRIs, the largest category of antidepressants, 30% to 45% have no clinical response. In addition, current therapies for the treatment of schizophrenia and depression do not address the cognitive impairments associated with these disorders. In addition to efficacy limitations, the majority of patients taking currently approved drugs for the treatment of Alzheimer’s disease, schizophrenia and depression experience one or more side effects. These can include, depending on the drug, headaches, nausea, nervousness, insomnia, agitation, weight gain and sexual dysfunction. Side effects occur because the biological mechanisms targeted by these drugs are not only involved in the neuronal pathways implicated in the disease but are also involved in multiple other bodily functions. These side effects can be so severe at higher dosage levels that in order to make the drug tolerable, the actual doses administered must be moderated. This, in turn, reduces the effectiveness of the treatment.

3

We believe that therapies designed to address biological targets within the cellular pathways critically involved in memory formation and other cognitive functions could be effective in treating many major neurological and psychiatric disorders. We also believe that due to the relative specificity of these targets, there are opportunities to reduce side effects.

THE MEMORY PHARMACEUTICALS DRUG DEVELOPMENT APPROACH

We are focused on the discovery and development of CNS drugs that impact biological targets believed to play a critical role in memory formation and cognition. Based on our understanding of neuronal pathways, we are targeting specific subcategories of neurotransmitter receptors, specific enzymes that regulate second messengers and protein channels that modulate the flow of second messengers such as calcium. We screen, optimize and develop highly selective compounds for these targets. Through this approach, we believe we can address many of the debilitating symptoms of several neurological and psychiatric disorders and potentially slow or halt the progression of some of these diseases. Compared with current alternatives for the treatment of Alzheimer’s disease, schizophrenia and depression, we believe that drug candidates we develop may provide enhanced efficacy and may have an improved side effect profile.

Our interdisciplinary drug discovery and development approach enables us to generate drug development programs focused on multiple targets and clinical indications. Through this approach, we believe that we are able to identify and optimize promising drug candidates quickly and efficiently. To accomplish this:

Ø	We use our extensive knowledge of the neuronal pathways involved in memory formation to identify highly relevant targets in the cascade of events leading to long-term memory formation.
Ø	We have created a dynamic, interdisciplinary environment in which experts in neuroscience, molecular biology, medicinal chemistry and preclinical drug development work closely together to screen, identify and optimize drug candidates.
Ø	We conduct iterative tests on our compounds to determine their pharmacokinetics, including their ability to be absorbed, distributed and metabolized in the body and their toxicity levels. In addition, we measure their biochemical and physiological effects in the body and their mechanisms of action, such as their ability to penetrate the blood-brain barrier.
Ø	Beginning at the earliest stages of drug discovery and continuing throughout the drug development process, we screen our compounds to eliminate those that are not specific to an identified target and therefore either exhibit increased side effects or cannot be optimized for a particular disease. In this process we use many neurobehavioral animal models and other sophisticated screening tools. We believe that we apply these screening techniques earlier in the development process than most other biopharmaceutical companies.
Ø	We complement our scientific disciplines with leadership by individuals who have extensive expertise and experience in drug discovery and development, related technologies and collaboration management.

We design our drug discovery process with the goal of significantly enhancing our ability to develop high quality drug candidates on an accelerated basis and to reduce our risk of failure in the clinical trial process. We believe that this approach helped us to secure our two collaborations with Roche, each of which was entered into following preclinical tests of our compounds. As of December 31, 2004, we had received $45.4 million in upfront license fees, payments for research and development services, milestone payments and an equity investment under these collaborations. In addition to royalties on any product sales, we could receive up to $248.0 million from these collaborations (based on achievement of all of the milestones for products for each of the indications specified in the collaboration agreements and including upfront license fees).

THE MEMORY PHARMACEUTICALS STRATEGY

Our goal is to be a leading biopharmaceutical company focused on the discovery and development of drug candidates for the treatment of CNS disorders that exhibit significant impairment of memory and other cognitive functions. In order to achieve this goal, we:

Ø

Aggressively pursue the discovery and development of drug candidates in large disease markets in which there are significant unmet medical needs. We are initially focused on the discovery and development of drug candidates for major CNS diseases for which the market opportunities are extensive and current therapies generally provide limited effectiveness and undesirable side effects. These diseases include Alzheimer’s disease, schizophrenia and depression. We intend to apply our expertise in the future to encompass a wider spectrum of neurological and psychiatric disorders by expanding the application of existing compounds and optimizing new compounds for these and other disorders such as vascular dementia, MCI and Parkinson’s disease, age associated cognitive decline, memory impairment from coronary artery bypass grafting, attention deficit hyperactivity disorder, chemotherapy-induced cognitive impairment, obsessive compulsive disorder and alcohol-induced cognitive impairment.

4

Ø	Leverage our extensive understanding of neuronal pathways and our interdisciplinary, accelerated drug development approach to yield multiple promising drug candidates. We use our knowledge of the critical pathways involved in memory formation to identify multiple targets that we believe are highly relevant to many different CNS disorders. We optimize compounds to act on targets that affect cellular pathways that are involved in more than one CNS disorder and we seek to develop multiple compounds to act on multiple targets in the same disorder. For example, for the treatment of Alzheimer’s disease, we are engaged in preclinical or clinical testing for two compounds that regulate cAMP, one compound that regulates neuronal calcium and one compound that acts as a partial agonist of the nicotinic alpha-7 receptor. In addition, one of these compounds regulating cAMP is also being evaluated as a potential candidate for the treatment of depression. Compounds in our PDE10A inhibitor program target the regulation of cAMP and cGMP and may be useful in the treatment of schizophrenia, Parkinson’s disease and obsessive compulsive disorder.
Ø	Establish collaborations with leading pharmaceutical and biotechnology companies to advance our drug candidates through clinical development and commercialization. We seek to leverage our pipeline of early-stage development candidates through collaborations with leading pharmaceutical and biotechnology companies. We believe that our focus on drug candidates with large market opportunities helps us secure these types of collaborations. We also believe that our rigorous approach to preclinical testing enables us to partner our compounds relatively early in the development process and on terms more in line with those achieved at later stages of clinical development. We seek to retain options to commercialize our compounds in certain territories or applications. We have a collaboration with Roche for our PDE4 inhibitors and a second collaboration with Roche for our nicotinic alpha-7 partial agonists. We are exploring the potential for a collaboration for MEM 1003 and for our PDE10A inhibitor program.
Ø	Seek to selectively in-license or acquire additional compounds. We seek to selectively augment our internal drug discovery program through the in-licensing or acquisition of advanced stage preclinical and clinical development stage compounds that are relevant to the pathways involved in memory formation and cognition. We believe that our expertise in these pathways enables us to identify promising compounds that can be converted to promising CNS drug candidates.

OUR DRUG DEVELOPMENT PROGRAMS

Our drug development pipeline currently includes four programs in clinical and preclinical stages (MEM 1003, PDE4 inhibitors, Nicotinic alpha-7 partial agonists and PDE10A inhibitors), and several other compounds with potential in our early stage pipeline. All of our drug candidates are small molecules designed for oral dosing. Our compounds that we have designated as drug candidates are as follows:

5

MEM 1003 for the treatment of Alzheimer’s disease, vascular dementia and MCI

Overview of Alzheimer’s disease

Alzheimer’s disease is a progressive neurodegenerative disorder for which there is no known cure. Later stages of Alzheimer’s disease involve severe loss of memory and other cognitive functions, including loss of awareness of recent experiences and of surroundings, and the inability to perform basic tasks. Patients with Alzheimer’s disease also suffer from significant personality and behavioral changes such as paranoia, delusions, hallucinations and compulsive, repetitive behaviors.

The exact cause of Alzheimer’s disease is currently unknown. All Alzheimer’s patients, however, exhibit a variety of characteristic pathological abnormalities. One characteristic of Alzheimer’s disease pathology is the presence of amyloid plaques and neurofibrillary tangles. The plaques are clumps of proteins that form deposits around neurons, while the tangles are clumps of proteins that accumulate inside neurons. Plaques and tangles are thought to interfere with normal neurotransmitter function and to have a toxic effect on neurons. Preclinical data from several laboratories suggests that the amyloid protein from which plaques are formed disrupts the function of the nicotinic alpha-7 receptor, which modulates neurotransmitter activity known to be required for the formation of long-term memories and possibly other cognitive functions as well.

Other pathological abnormalities present in Alzheimer’s patients involve dysfunction of neurotransmitters. Alzheimer’s patients exhibit a deficiency of the neurotransmitter acetylcholine, or ACh. This neurotransmitter is involved in short-term memory, alertness, and attention, and consequently reduced ACh levels are thought to be associated with cognitive impairment. In addition, post-mortem studies of Alzheimer’s patients have identified changes in the neurotransmitter glutamate, suggesting that over-stimulation of glutamate is in part responsible for neuronal degeneration.

Alzheimer’s patients also exhibit characteristic neuronal imbalances in the second messengers calcium and cAMP. Clinical research indicates that cAMP plays a role in memory, suggesting that drugs directed at this target will have beneficial effects on the treatment of memory loss associated with Alzheimer’s disease. One of the earliest manifestations of Alzheimer’s disease is an impaired regulation of calcium within CNS neurons. While small amounts of calcium are essential for memory formation and other cognitive functions, too much calcium causes a wide variety of toxic symptoms. Neuronal calcium levels are regulated by specific proteins known as L-type calcium channels. Abnormal regulation of these channels is believed to be an early step in the Alzheimer’s disease process, first impairing the pathways required for memory and other cognitive functions and eventually causing the death of the neurons.

Continuing studies of Alzheimer’s disease suggest that it is a highly complex disease that implicates numerous neurobiological pathways, presenting multiple opportunities for therapeutic intervention.

Current therapies for the treatment of Alzheimer’s disease

The principal treatments currently approved for Alzheimer’s disease belong to a class of drugs called acetylcholinesterase inhibitors. Due to the relationship between ACh and Alzheimer’s disease, it was originally thought that addressing the ACh deficit would have a significant therapeutic impact on Alzheimer’s disease. Acetylcholinesterase inhibitors are designed to impede the breakdown of ACh by an enzyme called acetylcholinesterase and work to keep levels of ACh high, although the cells that produce ACh continue to become damaged or die. Despite the relationship between ACh and Alzheimer’s disease, acetylcholinesterase inhibitors are not particularly effective in either treating the symptoms or stopping the progression of the disease. Only about 50% of patients taking acetylcholinesterase inhibitors for the treatment of Alzheimer’s disease experience even a modest improvement in symptoms. For those who show improvements, these benefits typically last for only six to nine months.

Acetylcholinesterase inhibitors produce a variety of side effects, including nausea, vomiting, loss of appetite, diarrhea and muscle cramps. These side effects significantly limit the drug dosage that can be safely administered, thereby restricting the ability to dose at levels that might further slow the deterioration of neurons and impact the progression of the disease.

A second category of drugs, NMDA-receptor antagonists, is designed to mitigate the impact of excessive glutamate release by blocking N-methyl-D-aspartate, or NMDA, receptors. One drug in this class has recently been approved in the US for treatment of moderate to severe cases of Alzheimer’s disease.

There have also been numerous attempts to develop drugs for Alzheimer’s disease that directly target the characteristic plaques and tangles. While many of these drug development programs have been abandoned, there are still programs that continue to focus on this method of action, though no drug candidates from these programs have been approved in the US.

MEM 1003

MEM 1003 is a neuronal L-type calcium channel modulator that we are developing for the treatment of Alzheimer’s disease. By blocking L-type calcium channels, MEM 1003 may regulate the flow of calcium. We believe that this approach may enhance

6

cognition by re-establishing normal levels of calcium, which is essential for normal functioning of the neuronal pathways. Further, we believe that this modulation of the calcium flow may impact the progression of the disease by protecting these neurons from further damage caused by the amyloid plaques.

We have conducted preclinical tests of MEM 1003 in vitro as well as in animal models. In our in vitro tests, MEM 1003 significantly blocked and regulated the function of neuronal L-type calcium channels in brain slices and single cell cultures.

We have conducted animal tests of MEM 1003 to assess its effect on learning and memory and its toxicology profile. We used various behavior assays to test the effect of MEM 1003 on memory and learning in four different species of varying ages, including aged non-human primates. In each of these tests, MEM 1003 improved memory and learning in healthy and cognitively impaired animals. MEM 1003 achieved these results using oral administration across a wide dose range. We have also tested the toxicology of MEM 1003 in two species over 13 weeks. There were no significant adverse toxicological findings in these studies, except as expected, at high dose levels.

In July 2003, we completed dose-escalating, double-blind, placebo-controlled Phase Ia and Ib trials with 125 healthy volunteers in the United Kingdom to evaluate the potential cardiovascular impact and side effect profile produced by single and multiple doses of MEM 1003. Over a dose range of up to 180 milligrams, which was the maximum dose tested, MEM 1003 was generally well tolerated. For doses of MEM 1003 up to the maximum tested, neither young nor elderly (over age 55) volunteers experienced any significant cardiovascular or other significant adverse side effects. In our human trials, the concentration of MEM 1003 in blood plasma exceeded the concentration levels which we achieved in our animal efficacy models. This leads us to believe that MEM 1003 may achieve the same therapeutic benefits in humans as it achieved in our animal models.

In addition to testing safety, we assessed cognitive enhancement in 40 of the volunteers from our Phase Ib trials. Both young and elderly healthy volunteers were tested with a widely used set of cognitive tests. In these tests, there were statistically significant effects on several of the cognitive measures of longer-term aspects of memory.

We are currently designing a Phase IIa randomized, multi-center, double-blind, placebo-controlled clinical study with this drug candidate, which we expect to commence by the end of the first half of 2005. In January 2005, we commenced a safety and tolerability study in patients with Alzheimer’s disease under a US Investigational New Drug (IND) application to assist us in assessing the differences in the safety and tolerability profile of MEM 1003 in healthy volunteers and Alzheimer’s patients in order to design the Phase IIa study. The safety and tolerability study is a single-center, randomized, double-blind, placebo-controlled clinical study. The study design consists of two segments, a double-blind dose escalation segment (the first segment) and double-blind treatment segment (the second segment). Approximately 85 patients are expected to be tested over the course of the two segments of the study. In the first segment, a certain dose level of MEM 1003 or placebo will be administered to patients twice on one day of certain weeks during the study period. Patients in this first segment will be treated at escalating doses of MEM 1003, which may reach up to 120 milligrams per dose. The highest tolerated dose of MEM 1003 in the first segment will then be administered to patients participating in the second segment of the study. In the second segment, successive groups of patients will be administered MEM 1003 or placebo twice daily for a period of ten days. In addition to measuring safety and tolerability, a secondary objective of the study is to evaluate the acute neurocognitive function in Alzheimer’s patients using a battery of standard cognitive tests. Pharmacokinetic measurements, or how the body absorbs, distributes, metabolises and eliminates MEM 1003, and other observations from the safety and tolerability study, including the results of the cognitive tests, will be used to select the appropriate dose(s) and finalize design of the Phase IIa clinical study.

The overall timing of the safety and tolerability study has been extended as a result of slower than expected patient enrollment and our decision to increase the number of patients being studied. While we currently expect to commence our Phase IIa study by the end of the first half of 2005, our ability to do so will depend upon our achieving patient enrollment sufficient not only to complete the safety and tolerability study but also to commence the Phase IIa study in the time frame that we have planned.

We are funding the safety and tolerability study and intend to fund the Phase IIa clinical trials for MEM 1003 ourselves. However, we are also exploring the potential for a collaboration for MEM 1003.

Similar to Alzheimer’s disease, patients with vascular dementia and MCI are also believed to exhibit an inability to regulate the level of calcium in the CNS. Due to this common defect, we believe that MEM 1003 has the potential to be effective in vascular dementia and MCI. We are currently focusing on clinical trials for MEM 1003 in Alzheimer’s disease, and depending on the outcome of this program, we may pursue MEM 1003 for other indications, including vascular dementia and MCI in the future.

7

MEM 1414 for the treatment of Alzheimer’s disease and potentially MCI

MEM 1414 is a PDE4 inhibitor and our second drug candidate for the treatment of Alzheimer’s disease and potentially for MCI. PDE4 inhibitors are designed to inhibit the activity of PDE4, the enzyme which breaks down neuronal cAMP. Studies have shown that administering PDE4 inhibitors can have a restorative effect on memory loss in animal models, including those of Alzheimer’s disease. Based on these findings, in addition to our own research on the neuronal pathways involved in Alzheimer’s disease, we believe that we may be able to treat this disease by increasing the levels of neuronal cAMP.

In preclinical tests of cognition in several species, MEM 1414 reversed memory deficits and demonstrated significant and sustained improvements in cognitive function over a wide dose range. Furthermore, after a single dose applied to a rat model, positive effects on long-term memory were sustained over a three-month period.

Under our 2002 Roche collaboration, Roche has exclusive rights to MEM 1414 and is currently conducting Phase I clinical trials. In a single ascending dose study in which groups of healthy volunteers were given a single dose of MEM 1414 at a particular dose level for each group, MEM 1414 was generally well tolerated across a wide dose range. A multiple ascending dose Phase I clinical trial, in which each participating group of healthy volunteers is receiving multiple doses at a particular dose level for each group, is currently underway.

Although the single doses of MEM 1414 administered in the Phase I single ascending dose study were generally well tolerated even at the highest doses administered, in subsequent preclinical animal tests conducted by Roche, MEM 1414 has caused adverse side effects, which in one study resulted in deaths at high doses in one of two animal species tested. As a result, Roche initiated additional investigational animal toxicology and other assessments of MEM 1414, and reduced, or capped, the maximum dosage to be administered in the multiple ascending dose Phase I clinical trial from the originally established dose range. The maximum dose in the multiple ascending dose Phase I clinical trial was subsequently raised beyond the initial cap, though analysis of the additional animal toxicology and other assessments of MEM 1414 are ongoing. The results of the current clinical trials and the animal studies could impact the continuation of the MEM 1414 program under the Roche collaboration, as well as potentially affect Roche’s development plans for MEM 1917, the back-up candidate to MEM 1414 for the treatment of Alzheimer’s disease.

Subsequent animal studies and clinical trials may fail to support our earlier preclinical and clinical trial results, particularly given that PDE4 inhibitors have generally exhibited safety and tolerability problems that have limited the ability of these compounds to be used as effective drugs.

MEM 1414 may also have applicability for other indications, including MCI.

MEM 3454 for the treatment of schizophrenia

Overview of schizophrenia

Schizophrenia is a neurological brain disorder that interferes with a person’s ability to think clearly, manage emotions, make decisions, and relate to others. Specific symptoms include delusions and hallucinations, an inability to sort and interpret incoming sensations and a corresponding inability to respond appropriately, and an altered sense of self. No single defect appears to be responsible for schizophrenia. Rather, multiple genetic and environmental factors seem to contribute to disturbances in brain function.

Current therapies for the treatment of schizophrenia

Current drug treatments for schizophrenia modulate primarily the level of the neurotransmitter dopamine by blocking the actions of this substance at the receptor level. This approach is based on the belief that decreasing the activity of the receptor negates the impact of the increase in the neurotransmitter. Other drug therapies target multiple neurotransmitter systems and receptors. Approximately 30% of individuals with schizophrenia do not respond to current medications at all and, even for those who benefit, these drugs often do not address the social withdrawal, affective changes and emotional behavior associated with the disease. Moreover, existing drugs are not designed to address or resolve the cognitive impairments associated with this disease. Current drug therapies for schizophrenia produce a variety of side effects. In particular, some of these antipsychotic drugs may result in weight gain, which can lead to an increase in the incidence of Type II diabetes and stroke.

MEM 3454

MEM 3454 is a partial agonist of the nicotinic alpha-7 receptor that we are developing as a potential therapy for schizophrenia. Studies have shown that schizophrenic patients can experience relief of symptoms through the administration of nicotine, which binds to nicotinic receptors. We believe that these nicotinic receptors play a significant role in long-term memory formation and other

8

cognitive functions. Nicotine, however, is associated with significant side effects, including cardiovascular disease, and cannot effectively be used as a therapeutic agent.

One of the nicotinic receptors is the nicotinic alpha-7 receptor. The nicotinic alpha-7 receptor is a highly specialized receptor found in the CNS, particularly in the hippocampal region of the brain. This receptor modulates neurotransmitter activity known to be required for the formation of long-term memories, and we believe that it plays an important role in other cognitive functions as well. The nicotinic alpha-7 receptor has been linked genetically with schizophrenia. We believe that compounds acting on the nicotinic receptor could be beneficial in the treatment of schizophrenia.

An agonist is a molecule that binds to a receptor and mimics the action of the transmitter specific for that receptor. Full agonists tend to lose their ability to induce activity after repeated administration. Partial agonists, on the other hand, do not typically result in the same desensitization. We have combined our knowledge of receptor action in the brain with our understanding of nicotinic alpha-7 receptors to develop partial agonists for this receptor. MEM 3454 is the lead compound from our nicotinic alpha-7 partial agonists series.

MEM 3454 has demonstrated efficacy in several preclinical animal models of schizophrenia. MEM 3454 generated positive results in preclinical models of antipsychotic drug action, including the animals’ ability to process information in a coordinated fashion. MEM 3454 also demonstrated activity in multiple animal models of cognition. In addition, in a preclinical test administering MEM 3454 for seven days, there was no evidence of weight gain in the animal models.

In February 2005, we commenced a Phase I clinical trial of MEM 3454 in healthy volunteers under a Clinical Trial Application that we filed with Health Canada. The study is a single-center, double-blind, placebo-controlled clinical study to evaluate the safety, tolerability and pharmacokinetics of single ascending doses of MEM 3454.

Under our 2003 Roche collaboration, Roche has the right to obtain an exclusive license for MEM 3454 following the completion of Phase IIa clinical trials by making payments to us upon our achievement of certain developmental milestones. Alternatively, Roche has the right to terminate its license option with respect to MEM 3454 or any other product covered by that agreement on a product-by-product basis at certain times during the development of the product. If at any point during the development of MEM 3454, Roche chose not to maintain its license option with respect to this drug candidate, we would not receive further milestone payments or other support from Roche for its continued development.

MEM 1917 for the treatment of Alzheimer’s disease and depression

MEM 1917 is a PDE4 inhibitor which is being evaluated by Roche as a back-up candidate to MEM 1414 for the treatment of Alzheimer’s disease. As is the case with MEM 1414, MEM 1917 is designed to inhibit the activity of PDE4, the enzyme that breaks down neuronal cAMP.

MEM 1917 is also being evaluated by Roche as a potential candidate for the treatment of depression. We believe that the pathways involved in memory formation may also be implicated in depression, although this link is not fully established or understood. If there are shared pathways in memory formation and depression, increasing levels of cAMP by inhibiting the action of PDE4 may remediate the neuronal imbalances and the cognitive impairments that occur in depressed patients.

Selective serotonin reuptake inhibitors and dual uptake inhibitors are the principal drugs currently approved for depression. These drugs work by impeding the brain’s breakdown of serotonin, a key neurotransmitter that is depleted in depressed patients and, in the case of dual uptake inhibitors, a second neurotransmitter norepinephrine as well. These existing drugs can produce several side effects, including headaches, nausea, nervousness, insomnia, agitation and sexual dysfunction, which significantly limit the dosage that can be safely administered. This, in turn, can reduce the effectiveness of these drugs. Another major drawback of these therapies is their relatively slow onset, requiring several weeks of administration before a patient may experience a therapeutic effect. We believe that PDE4 inhibitors, through their impact on cAMP, may provide the potential for a safer and faster-acting therapy than current therapies for depression.

In multiple animal behavior models of cognitive impairment, MEM 1917 has reversed memory deficits. In addition, in several animal models of depression, MEM 1917 has demonstrated efficacy in both short-term and long-term tests of antidepressant activity. Both Roche and we have conducted toxicology studies and are currently conducting preclinical animal studies for efficacy in Alzheimer’s disease, cognition and depression. The results of these studies, together with the comparative profile of MEM 1917 relative to MEM 1414, will be considered by Roche in determining the development plans for this compound.

Under our 2002 Roche collaboration, Roche has an exclusive license for MEM 1917.

9

MEM 63908 for the treatment of Alzheimer’s Disease

MEM 63908 is a partial agonist of the nicotinic alpha-7 receptor and the fourth drug candidate that we are developing as a potential treatment for Alzheimer’s disease.

Among other pathological abnormalities, Alzheimer’s patients typically exhibit amyloid plaques, which are clumps of protein that form deposits around neurons. Preclinical data from several laboratories suggests that the amyloid protein from which these plaques are formed disrupts the function of the nicotinic alpha-7 receptor. We believe that as a partial agonist of the nicotinic alpha-7 receptor, MEM 63908 could overcome the effect of the amyloid protein and thereby be beneficial in the treatment of Alzheimer’s disease.

We have conducted animal tests of MEM 63908 to assess its effect on learning and memory. We used various behaviour assays to test the effect of MEM 63908 on memory and learning. In these tests, MEM 63908 improved learning and memory in healthy and aged cognitively impaired animals. MEM 63908 has also demonstrated a good safety profile in toxicological and safety studies.

Under our 2003 Roche collaboration, Roche has the right to obtain an exclusive license for MEM 63908 following the completion of Phase IIa clinical trials by making payments to us upon our achievement of certain developmental milestones. Alternatively, Roche has the right to terminate its license option with respect to MEM 63908 or any other product covered by that agreement on a product-by-product basis at certain times during the development of the product. If at any point during the development of MEM 63908, Roche chose not to maintain its license option with respect to this drug candidate, we would not receive further milestone payments or other support from Roche for its continued development.

PDE10A Inhibitor Program and other drug discovery programs

PDE10A is an enzyme that has been shown to be present at high levels in neurons in areas of the brain that are closely associated with many neurological and psychiatric disorders such as schizophrenia, Parkinson’s disease and obsessive compulsive disorder. PDE10A degrades the intracellular signaling molecules cAMP and cGMP, molecules that are responsible for improving the function of many different cells in the body, including neurons. By inhibiting PDE10A, levels of cAMP and cGMP are increased within neurons and the ability of these neurons to function properly is thereby improved. We believe that PDE10A inhibitors may be useful in treating a range of neurological and psychiatric disorders.

We have identified two chemical series of compounds that have demonstrated effectiveness in the selective inhibition of PDE10A activity. Representatives of these PDE10A inhibitors have been shown to be effective in several animal models of schizophrenia. We are in the process of identifying potential drug candidates from these series.

We are exploring the potential for a collaboration for our PDE10A inhibitor program.

In addition to our preclinical and clinical drug candidates, we continue to work on new targets and new chemistries. We are testing a number of compounds in vitro and in animal models for efficacy, safety and tolerability. Our pipeline includes additional compounds that we are exploring for the treatment of Alzheimer’s disease, schizophrenia, depression, vascular dementia and MCI. In addition, we are working on compounds that may have potential application to other CNS disorders, such as Parkinson’s disease and age associated cognitive decline.

RESEARCH AND DEVELOPMENT

Our drug discovery programs are designed to yield effective and safe drug candidates in a shorter timeframe relative to industry norms. By performing benchmark safety and efficacy tests early in the process and only advancing the most promising drug candidates, we believe that we are able to accelerate the overall drug development timeline and provide higher quality drug candidates with reduced risk of failure in clinical trials. We strive to achieve a high level of productivity in our drug discovery and development programs in part through the close integration of our research group and development group staff.

Our research efforts focus on identifying new molecular targets that play a role in the formation, modification and stabilization of synaptic connections in the human brain. We clone these targets and study their functions by developing and applying various assays. To discover active compounds, we screen the targets against libraries of compounds that embody properties that would make them suitable for use as CNS drugs. Compounds are tested and optimized in an iterative process designed to produce highly specific, orally available drug candidates suitable for CNS indications. Set forth below is a summary description of the five principal components of our drug selection process.

Target selection

We select targets that we believe have an important role within the pathways involved in long-term memory formation, are well distributed in the relevant areas of the brain, and are within the class of proteins that are suitable as drug targets. These targets may regulate neurotransmitter signaling, modulate second messengers that carry the signal from the synapse to the rest of the neuron, or sensitize the receptors and channels in the synapse.

10

CNS chemistry

Using internal and external chemistry resources, we employ a combination of medicinal, combinatorial and analytical chemistry procedures to maximize the chances of discovering specific and efficacious lead compounds. To produce lead optimization candidates, we employ an iterative process in which medicinal chemists who synthesize novel chemical entities work with our neurobiologists and pharmacologists to improve the desirable features of the compound, while reducing the potential for undesired characteristics. We select compounds strategically, and synthesize focused chemical libraries based on their degree of target specificity, binding properties and biological functionality.

Animal model screening processes

We have developed a series of assays that act within progressively more complex and realistic biological environments in order to systematically evaluate potential drug candidates for their effect on memory and other cognitive functions. The process begins with cell-based assays performed in test tubes, which assist us in understanding the biochemical effect of compounds. Next, proprietary assays are used to measure electrical activity in brain slices of rats and mice. These sophisticated neurobehavioral assays allow us to determine the neurophysiological effect of compounds before using more complex and expensive animal models. Finally, we have developed animal behavior models, using both young and old rats and mice that allow us to assess different aspects and phases of long-term memory and cognition. We use naturally-aged impaired mice and rats for these studies, which we believe ensures a more realistic physiology than would be available in young animals with artificially-induced age characteristics. We believe that the animal models we have developed and employ enhance our ability to predict promising drug candidates on an accelerated and more focused and informed basis.

Preclinical safety and pharmacokinetics

Our safety specialists study the interaction of our potential drug candidates with the intended target as well as with the rest of the body. They measure the ability of the compound to be absorbed and to reach the intended target, and they determine how the drug is metabolized within, and removed from, the body. They also study the potential for side effects.

An anatomical and functional feature of the brain that makes it difficult to formulate CNS-specific therapeutics is the blood-brain barrier. This barrier can prevent drugs from entering the brain from blood vessels. We have significant expertise and experience in using rat and mouse models to assess the plasma/brain ratios of compounds in a steady state, which is indicative of a compound’s ability to penetrate the blood-brain barrier. We control these analytical tests by conducting these analyses in our own laboratories, which are designed to meet FDA Good Laboratory Practice standards.

Bioinformatics and information technology systems

We deploy an advanced informatics infrastructure to support our research and development efforts. Our hardware infrastructure includes a high-speed internal network with data integrity and security measures. Our software infrastructure is tailored to address specific needs, including research logistics, scientific data management and analysis. Productivity-enhancing systems have been put in place to handle many routine aspects of conducting pharmaceutical research. An enterprise-wide scientific data management system captures critical information on chemical structures and associated biological data to create a shared knowledge-base designed to accelerate the drug development process. We employ scientists cross-trained in information technology to assess and implement commercially available scientific computing tools or to develop new tools internally.

COLLABORATIONS AND IN-LICENSES

We seek to enter into collaborations with leading pharmaceutical and biotechnology companies and we believe that this enables us to leverage these companies’ resources to exploit our drug candidates on a global basis, while allowing us to remain focused on early stage development and discovery of drug candidates. To date, we have established two major collaborations, both of which are with Roche. One of the collaborations is for the development of our PDE4 inhibitors, including MEM 1414 and MEM 1917, and the other collaboration is for the development of our nicotinic alpha-7 partial agonists, including MEM 3454 and MEM 63908.

Another important component of our strategy is to augment our internal drug discovery programs through the selective in-licensing of advanced-stage preclinical and early-stage clinical chemical compounds that we believe are relevant to the pathways involved in memory formation and other cognitive functions.

11

2002 Roche collaboration

In July 2002, we entered into a collaboration with Roche for the development of PDE4 inhibitors for neurological and psychiatric indications, and potentially other indications, which we extended and expanded in August 2004. We granted Roche a worldwide, exclusive, sub-licensable license to our patent rights and know-how with respect to any PDE4 inhibitor for the prevention and treatment of diseases, in all indications, for either human or veterinary use. If Roche decides to co-promote any product covered by the agreement in the US, Roche must offer us this co-promotion opportunity before it may offer a right to co-promote such product to any non-affiliated third party. Roche is obligated to provide, at its own expense, clinical supplies of compounds covered by this agreement.

Under the 2002 Roche collaboration, Roche has paid us a total of $21.6 million through December 31, 2004, comprised of an upfront license fee, payments for research and development services and one milestone payment. Roche is obligated to make milestone payments to us if we achieve specified development, regulatory and commercialization milestones (including sales level milestones following a product’s launch) for certain indications. These indications include neurological, psychiatric and non-CNS indications for our PDE4 inhibitors, as well as any neurological or psychiatric indication for any PDE4 inhibitor developed by Roche without using our technology. If we reach all of the milestones for one neurological indication and one psychiatric indication for our PDE4 inhibitors under the agreement, milestone payments payable to us under the agreement, including the upfront license fee and bonus payments for achieving certain sales milestones, could total $106.0 million. In addition, we are entitled to receive up to $34.5 million in milestone payments if we develop compounds for both a second neurological indication and an indication other than a neurological or psychiatric indication. We are also entitled to royalties based on a specified percentage of net sales of products, which increases at increasing net sales levels, during the term of the agreement. Royalty payments will expire generally on a country-by-country basis on the later to occur of (i) the expiration of the last to expire patent containing a composition of matter claim in a given country and (ii) ten years following the launch of the product in that country.

The agreement may be terminated by either party following an uncured material breach by the other party. In addition, we have the right to terminate the agreement on a product-by-product or region-by-region basis, or in its entirety in certain cases, if Roche does not use reasonable diligence to develop, obtain regulatory approvals for, manufacture or commercialize products in certain key countries. On August 6, 2004, the research collaboration under this agreement with Roche was extended and expanded for a two-year period. In connection with this extension, Roche has committed to a minimum of 18 months’ additional funding of our research collaboration efforts in the aggregate amount of $5.25 million, payable quarterly commencing September 9, 2004. At any time after September 9, 2005, Roche has the right to terminate the research collaboration or the agreement on a region-by-region or product-by-product basis or in its entirety, upon the giving of six months advance notice, or, in the case of the agreement, without notice in the event that we do not consent to a proposed sublicense by Roche to a third party. Unless the research collaboration is earlier terminated by Roche, our research collaboration efforts under this agreement will be funded by Roche through September 9, 2006. Unless it is earlier terminated, the agreement will terminate on the date of expiration of all royalty and other payment obligations under the agreement.

Upon the termination of the agreement as to any country, region or product, the rights and licenses granted to Roche under the agreement as to such country, region or product will terminate, except that Roche can use collaboration technology other than for PDE4 inhibitors. In addition, at our request, Roche is required to transfer certain filings, rights, approvals, agreements and data relating to such product to us at no expense. Further, upon any such termination, we would have a non-royalty bearing license under related Roche intellectual property rights to commercialize such product or in such country or region.

MEM 1414, which was accepted for the Roche development pipeline in March 2003, is included in this collaboration and in the license rights that we granted to Roche. In August 2003, MEM 1414 entered Phase I clinical testing, with initial development focusing on the treatment of Alzheimer’s disease. MEM 1917, a backup candidate to MEM 1414 for the treatment of Alzheimer’s disease and a potential candidate for the treatment of depression, is also licensed to Roche under this collaboration.

2003 Roche collaboration

In August 2003, we entered into a second collaboration with Roche to develop nicotinic alpha-7 partial agonists for treating CNS indications including schizophrenia and Alzheimer’s disease. We granted to Roche the right, on a compound-by-compound basis, to obtain an exclusive, worldwide, sub-licensable license to our patent rights and know-how for any nicotinic alpha-7 partial agonist that we develop, during the five years following the commencement of the collaboration, for the prevention or treatment of any diseases in any indication, for either human or veterinary use. This right is exercisable by Roche upon completion of Phase IIa clinical trials for a covered compound, or earlier under certain circumstances, and allows us to continue to manage at our expense the development of our compounds covered by the collaboration through the completion of Phase IIa clinical studies. We have the right to co-promote any product covered by the agreement in the US, on a product-by-product basis, subject to the terms of a separate co-promotion agreement, if we pay specified percentages of Roche’s Phase III global development costs for that product. If we exercise our co-promotion right with respect to a particular product, we will be entitled to a specified percentage of the gross profits from US sales of

12

that product, in exchange for a reduction in the royalties on US sales otherwise payable to us by Roche. Upon exercise by Roche of its right to license a compound, Roche is obligated to provide, at its own expense, clinical supplies of such compound.

Under the 2003 Roche collaboration, Roche has paid us a total of $23.8 million through December 31, 2004, comprised of an upfront license fee, payments for research and development services and an equity investment. Under the 2003 Roche collaboration, Roche paid us an upfront license fee of $10.0 million in September 2003 and is obligated to pay us $6.0 million for research and development services over the first two years of the agreement. If Roche maintains its right to the license and exercises the right, Roche is obligated to make milestone payments to us if we achieve specified development, regulatory and commercialization milestones (including sales level milestones following a product’s launch) for certain indications. These indications include neurological, psychiatric and non-CNS indications for compounds developed by us or as part of the collaboration. If we reach all of the milestones for one neurological indication and one psychiatric indication for compounds developed under the agreement, milestone payments payable to us under the agreement, including the upfront license fee and bonus payments for achieving certain sales milestones, could total $142.0 million. In addition, if we develop compounds for additional indications, we are entitled to receive milestone payments (i) up to $42 million for each compound for each additional neurological or psychiatric indication, depending on the indication and the timing of milestone achievement, and (ii) up to $13 million for each compound for any indication other than a neurological or psychiatric indication. We are also entitled to royalties based on a specified percentage of net product sales, which increases at increasing net sales levels, during the term of the agreement. Roche’s royalty payment obligations will expire generally on a country-by-country basis on the later to occur of (i) the expiration of the last to expire patent containing a composition of matter claim in a given country; and (ii) ten years following the launch of the product in that country. In connection with the collaboration, Roche purchased shares of our series Roche preferred stock for an aggregate of $10.0 million (which converted into 925,926 shares of common stock on the closing date of our initial public offering), and received warrants to purchase 115,740 shares of common stock at an exercise price of $12.96 per share.

The collaboration agreement may be terminated by either party following an uncured material breach by the other party. In addition, Roche has the option to maintain its license rights with respect to any product covered by the agreement on a product-by-product basis. If Roche chooses not to exercise its option to maintain its license rights with respect to a product, we would not receive further milestone payments or other support from Roche for the development of that product. Roche has the right to terminate the agreement on a product-by-product or region-by-region basis, or in its entirety in certain cases, if we do not use reasonable diligence in developing nicotinic alpha-7 partial agonists through Phase IIa clinical trials. We have the right to terminate the agreement on a product-by-product or region-by-region basis, or in its entirety in certain cases, if Roche does not use reasonable diligence in meeting its obligations to develop, obtain regulatory approvals for, manufacture and commercialize products in certain key countries. In addition, Roche has the right to terminate the agreement at any time, on a region-by-region or product-by-product basis or in its entirety, upon the giving of advance notice, or without notice in the event that we do not consent to a proposed sublicense by Roche to a third party in the US. Unless otherwise terminated, the agreement will terminate on the date of expiration of all royalty and other payment obligations under the agreement.

Upon termination of the agreement as to any country, region or product, the rights and licenses granted to Roche under the agreement as to such country, region or product will terminate. In addition, at our request, Roche is required to transfer certain filings, rights, approvals, agreements and data relating to such product to us at no expense. Further, upon any such termination, we have the right to obtain exclusive and royalty-bearing license rights under related Roche patent rights and know-how to commercialize such product in such country or region.

MEM 3454 and MEM 63908 are included in this collaboration and in the related license option rights that we have granted to Roche. MEM 3454 is in preclinical development for the treatment of schizophrenia and other indications and MEM 63908 is in preclinical development for the treatment of Alzheimer’s disease.

BAYER AG in-license

In June 2001, we entered into an agreement with Bayer AG, or Bayer, for an exclusive, worldwide, sub-licensable license under certain Bayer patents and know-how to MEM1003 for the treatment of human peripheral and CNS-related disorders. We have an option, exercisable until completion of the first Phase II trial for MEM 1003, to acquire a non-exclusive, world-wide license to certain Bayer patents and know-how relating to certain controlled release formulations developed for Bayer for an unrelated product. Under the agreement, we must use commercially reasonable efforts to develop products using the compound covered by this agreement and to commercialize those products in the US, Japan and four countries of the European Union to be selected by us. Bayer is under no obligation to provide us with any assistance in the development and commercialization of products covered by the agreement.

Under the license agreement, we made an upfront payment of $250,000 to Bayer as well as a milestone payment of $750,000 upon initiation of Phase I clinical trials. We are required to make additional payments of up to $19.0 million upon the achievement of specified milestones, including a $1.0 million milestone payment upon the commencement of Phase II clinical trials. We are also

13

obligated to pay royalties during the term of the agreement based on a specified percentage of worldwide net sales of products covered by the license agreement, which increases at increasing net sales levels and varies depending on whether the sales are made by us or by a sub-licensee. Royalty payment obligations will end on a country by country basis upon the expiration of the last patent rights in a given country or after ten years from the first commercial sale of a product covered by the license agreement in a given country, whichever is longer.

The license agreement may be terminated by either party following an uncured breach by the other party. Bayer also has the right to terminate the license agreement if we fail to make timely payments of any amounts due to Bayer under the license agreement. Unless otherwise terminated, the agreement terminates upon the expiration of the last-to-expire patent rights granted under the agreement. Issued patents from the US, Germany, France, Italy, Great Britain and Spain, the largest markets in which patents covered by the agreement have been issued, expire in late 2014 through 2015, although they are eligible for patent term extensions of up to five years. Patent applications are pending in Japan and Canada. Upon termination of the license agreement for any reason, other than an uncured breach by Bayer, all rights and licenses granted to us under the license agreement will terminate. In addition, upon termination of the license agreement as a result of our uncured breach, at Bayer’s request we will also have to disclose and transfer all of the rights we may have to the data and results, including any inventions, of our development efforts relating to MEM 1003.

INTELLECTUAL PROPERTY

Our success depends in part on our ability to obtain and maintain intellectual property protection for our drug candidates, technology and know-how, to operate without infringing the proprietary rights of others and to prevent others from infringing our proprietary rights. Our policy is to seek to protect our chemical compounds and technologies by, among other methods, filing US and foreign patent applications related to our proprietary technology, inventions and improvements that are important to the development of our business. We also rely on trade secrets, know-how, continuing technological innovation and in-licensing opportunities to develop and maintain our proprietary position. We, or our licensors, file patent applications directed to all drug candidates in an effort to establish intellectual property positions regarding new chemical entities relating to our product candidates as well as uses of new chemical entities in the treatment of CNS diseases. In total, as of March 10, 2005, we owned or licensed from Bayer seven issued US patents, 33 pending US patent applications, 110 issued foreign patents and 160 pending foreign patent applications.

The patent positions of companies like ours are generally uncertain and involve complex legal and factual questions. Our ability to maintain and solidify our proprietary position for our technology will depend on our success in obtaining effective patent claims and enforcing those claims once granted. We do not know whether any of our patent applications or those patent applications that we license will result in the issuance of any patents. Our issued patents and those that may issue in the future, or those licensed to us, may be challenged, invalidated, rendered unenforceable or circumvented, which could limit our ability to stop competitors from marketing related products or the length of term of patent protection that we may have for our products. In addition, the rights granted under any issued patents may not provide us with competitive advantages against competitors with similar compounds or technology. Furthermore, our competitors may independently develop similar technologies or duplicate any technology developed by us in a manner that does not infringe our patents or other intellectual property. Because of the extensive time required for development, testing and regulatory review of a potential product, it is possible that, before any of our drug candidates or those developed by our collaborators can be commercialized, any related patent may expire or remain in force for only a short period following commercialization, thereby reducing any advantage of the patent.

We may rely, in some circumstances, on trade secrets to protect our technology. However, trade secrets are difficult to establish and enforce. We seek to protect our proprietary technology and processes, in part, by confidentiality agreements with our employees, consultants, scientific advisors and other contractors. These agreements may be breached, and we may not have adequate remedies for any breach. In addition, our trade secrets may otherwise become known or be independently discovered by competitors. To the extent that our employees, consultants or contractors use technology or know-how owned by others in their work for us, disputes may arise as to the rights in related inventions.

MEM 1003

As a result of our in-licensing agreement with Bayer, we have rights to several patents and patent applications, including five issued US patents, one pending US patent application, 109 issued foreign patents, and 24 pending foreign patent applications relating to MEM 1003. Issued patents from the US, Germany, France, Italy, Great Britain and Spain, the largest markets in which patents covered by the Bayer agreement have been issued, expire in late 2014 through 2015, although they are eligible for patent term extensions of up to five years.

14

PDE4 inhibitors

We have two issued US patents, 17 pending patent applications in the US, one issued foreign patent, and 127 pending foreign patent applications for both chemistries and research tools relating to our PDE4 inhibitors, including MEM 1414 and MEM 1917. We have identified four different classes of chemical compounds that act as PDE4 inhibitors. One issued patent and several pending applications include claims relating to compounds as well as the methods of manufacturing and using them. One issued US patent and several pending patent applications are drawn to research tools relating to materials and methods useful for screening relating to PDE4.

Nicotinic alpha-7 partial agonists

We have 12 pending US patent applications and six pending foreign patent applications for both chemistries and research tools relating to our nicotinic alpha-7 partial agonists, including MEM 3454. Our patent applications relate to three classes of nicotinic alpha-7 partial agonists as well as related research tools including genes and methods useful for screening.

PDE10A inhibitor program

We have two pending US patent applications and one pending foreign patent application for both chemistries and research tools relating to our PDE10A inhibitors. Our patent applications relate to two classes of PDE10A inhibitors as well as related research tools including genes and methods useful for screening.

General research tools

We have one pending patent application in the US and one foreign patent application relating to our general research tools. These patent applications relate to research tools, including methods and genes, which are applicable to different programs.

Trademarks

Our trademarks are covered by various pending applications and registrations in the US Patent and Trademark Office and similar agencies in the European Union and certain other countries. We have applied for or received trademark protection as set forth below.

	Ø	We have a pending application in the US and registrations in Australia, the European Union and Mexico for MEMORY PHARMACEUTICALS TM .
	Ø	We have pending applications in the US, European Union, Canada, and Mexico, and registrations in Australia, Japan, Hong Kong and Mexico for the MEMORY design logo.
	Ø	We have a pending application for SCIENCE WORKING FOR THE MIND TM in the US.

COMPETITION

The development and commercialization of new drugs and drug delivery technologies is highly competitive. We and/or our collaborators will face competition with respect to any of our products developed or commercialized in the future from major pharmaceutical companies, specialty pharmaceutical companies and biotechnology companies worldwide. Many of our competitors possess greater financial, managerial, scientific and technical resources and have longer operating histories and established reputations for successfully developing and marketing drugs, all of which put us at a competitive disadvantage. We face and will continue to face competition in the discovery, in-licensing, development and commercialization of our drug candidates, which could severely impact our ability to generate revenue or achieve significant market acceptance of our drug candidates. Furthermore, new developments, including the development of other drugs, technologies and methods of preventing the incidence of disease, such as vaccines, occur in the pharmaceutical industry at a rapid pace. These developments may render our drug candidates or technologies obsolete or non-competitive.

Additionally, various large pharmaceutical companies, universities and public agencies are developing and using technologies to treat cognitive disorders directly as well as other neurological and psychiatric illnesses which may have the ancillary benefit of treating memory and other cognitive disorders. Many of these entities have significant experience in preclinical testing, human clinical trials, product manufacturing, marketing and distribution and the regulatory approval process. Many companies also have substantially greater resources and are developing or using technologies that may be competitive with our products and technologies.

There are several global pharmaceutical companies with drugs in the marketplace for the treatment of Alzheimer’s disease, schizophrenia and depression. Acetylcholinesterase inhibitors for the treatment of Alzheimer’s disease include Aricept (Pfizer),

15

Reminyl (Johnson and Johnson) and Exelon (Novartis). The only NMDA-receptor antagonist for the treatment of moderate to late stage Alzheimer’s disease in the US is Namenda (Forest Laboratories). The SSRI antidepressant market includes Prozac (Eli Lilly), Paxil (GlaxoSmithKline), Zoloft (Pfizer), Celexa (Forest Laboratories) and Lexapro (Forest Laboratories). The only approved dual uptake inhibitor for the treatment of depression is Effexor (Wyeth). The antipsychotics on the market for the treatment of schizophrenia include Zyprexa (Eli Lilly), Risperdal (Johnson and Johnson) and Abilify (Bristol-Myers Squibb).

We are aware that there are many drugs under development by both large pharmaceutical companies and small biotechnology companies for additional treatments of Alzheimer’s disease, schizophrenia and depression as well as potential therapies for vascular dementia and MCI.

We believe that our ability to successfully position ourselves within this competitive environment will depend on, among other things:

	Ø	efficacy, safety and reliability of our drug candidates;
	Ø	the speed at which we or our collaborators develop drug candidates;
	Ø	completion of clinical development and laboratory testing of our drug candidates;
	Ø	timing and scope of regulatory approval of our drug candidates;
	Ø	our or our collaborators’ ability to manufacture and sell commercial quantities of approved products in the market;
	Ø	product acceptance by physicians and other health care providers;
	Ø	skills of our employees and our ability to recruit and retain skilled employees;
	Ø	protection of our intellectual property; and
	Ø	availability of substantial capital resources to fund development and commercialization activities by us and our collaborators.

GOVERNMENT REGULATION

Government authorities in the US and in other countries extensively regulate, among other things, the research, development, testing, manufacture, labeling, promotion, advertising, distribution, marketing, and export and import of pharmaceutical products such as those we are developing. We cannot assure you that any of our drug candidates will prove to be safe or effective, will receive regulatory approvals or will be successfully commercialized.

US regulatory approval

In the US, drugs and drug testing are regulated by the FDA, as well as state and local government authorities. Before our products may be marketed in the US, we must comply with the Federal Food, Drug and Cosmetic Act, which generally involves the following:

	Ø	preclinical laboratory and animal tests performed under the FDA’s Good Laboratory Practices regulations (GLPs);
	Ø	submission and acceptance of an IND application, which must become effective before clinical trials may begin in the US;
	Ø	adequate and well-controlled human clinical trials to establish the safety and efficacy of the drug candidate in our intended use;
	Ø	development of manufacturing processes which conform to FDA-mandated current Good Manufacturing Practices (GMPs); and
	Ø	FDA review and approval of a New Drug Application (NDA) prior to any commercial sale or shipment of a product.

The testing and approval process requires substantial time, effort and financial resources, and we cannot be certain that any approval will be granted on a timely basis, if at all. In addition, final regulatory approval and/or the speed of regulatory approval may depend on conditions and events prevailing in the pharmaceutical industry that are outside our control.

Preclinical tests

Preclinical tests include laboratory evaluation of the drug candidate, its chemistry, formulation and stability, as well as animal studies to assess the potential safety and efficacy of the drug candidate. The results of the preclinical tests, together with manufacturing information, analytical data and other available information about the drug candidate, are submitted to the FDA as part of an IND. An IND i