Rubin Laboratory

Our laboratory focuses on translational, high-throughput research to model disease and develop drugs using stem cells. We have established an extensive range of complex image based assays that have been used to probe various properties of stem cells and of many cells derived from them. To accomplish this we use automated high content screening imagers, associated robotic equipment, and selected small molecule libraries. We combine our high throughput approach with detailed molecular studies to further our understanding of the mechanisms of disease and determine which compounds are most likely to be of therapeutic value.



A primary focus of the lab is diseases of motor neuron degeneration, including spinal muscular atrophy (SMA) and amyotrophic lateral scleroris (ALS).  SMA is genetic disorder caused by the decrease in the level of Survival of Motor Neuron protein (SMN) leading to neuromuscular degeneration in children. We have carried out quantitative high content measurements in patient fibroblasts to discover small molecules that increase the level of SMN. (Makhortova et al. Nature Chemical Biology, 2011). Prolonging motor neuron survival is a major therapeutic goal for treatment of both SMA and ALS. We have executed a variety of survival assays using motor neurons generated in large numbers from mouse and human ES cells as well as patient derived iPS cells to identify compounds that prevent motor neuron death.

Another important aim of our lab is to develop new stem cell based disease models using small molecule screens.  One project has applied this approach to improve the differentiation of stem cell derived dopaminergic neurons, the same cell type that is specifically lost in Parkinson’s disease. We are also working to identify small molecules that enhance transdifferentiation, which is the direct conversion of one differentiated cell type into another. 

Our expertise in developing stem cell based disease models and high content screening allows us to contribute to a large number of collaborations with members of the Harvard Stem Cell Institute.

Read more

Spinal muscular atrophy (SMA)

Spinal muscular atrophy (SMA) is an autosomal recessive disease caused by mutations in the Survival of Motor Neuron 1 (SMN1) gene. One of the primary features of SMA is the progressive loss of neuromuscular function that is often fatal, making SMA the leading genetic cause of death in infants and young children. Motor neuron death is the most significant feature of this disease, but muscle dysfunction or malformation may also occur.

While SMN protein appears to have multiple cellular roles, studies of SMA patient populations indicate that higher levels of SMN expression are associated with less severe disease. Mouse studies confirm this general finding. This suggests a clear therapeutic strategy: namely, identifying the pathways and, ultimately, drug classes that increase SMN levels. However, there are alternate strategies. One is the identification of pathways that function independently of SMN and are corrective when SMN levels are reduced. Another is the identification of pathways that enhance the function of the reduced amount of SMN found in SMA patients.

Research efforts in our lab are currently focused on all three of these approaches to provide insight into the biology of the SMN protein and the etiology of SMA. We are currently using motor neurons derived from mouse and human embryonic stem (ES) cells to screen for compounds or pathway modulators that elevate intracellular levels of SMN. Moreover, we are using these ES cell derived motor neurons to establish in vitro models of SMA which we are using to identify which of these compounds or pathway modulations result in a functional improvement. Other studies are also currently underway to determine the contribution that other cell types, such as muscle or glia, make to the etiology and progression of SMA. It is our hope that these studies may lead to the development of better therapies for this insidious disease.

Amyotrophic Lateral Sclerosis (ALS)

Amyotrophic Lateral Sclerosis (ALS) is a progressive and fatal neurodegenerative disease caused by the degeneration of both upper and lower motor neurons. The condition is often called Lou Gehrig's Disease in North America, after the New York Yankees baseball icon who was diagnosed with the disease in 1939 and died from it in 1941. Common clinical features of ALS include muscle weakness, fasciculations, brisk or depressed reflexes, and extensor plantar responses. Although motor deficit usually predominates in the limbs, bulbar enervation can be severely involved early in the course of the disease, leading to atrophy of the tongue, dysphagia, and dysarthria. 90% of people with ALS die from respiratory failure within three to five years from the onset of symptoms. ALS being the most common motor neuron disease in adults, there are currently no truly effective treatments for it. The only existing ALS drug, Riluzole, only extends life by months.

ALS is mostly sporadic in nature, but occasionally (~10%) the disease is inherited. Approximately 20% of familial ALS cases are linked to mutations in the superoxide dismutase (SOD1) gene. Most familial forms of ALS are clinically and pathologically almost indistinguishable from sporadic ALS. The phenotypic similarity between the genetic and sporadic forms of the disease indicates that they share important pathogenic mechanisms. Consequently, information obtained by studying ALS caused by mutant SOD1 should help understand the key cellular and molecular mechanisms of sporadic ALS. While the detailed way in which mutant SOD1 expression leads to motor neuron death remains elusive, cell autonomous (i.e. in motor neurons) and non-cell autonomous (e.g. through astrocytes and microglia) pathways appear to participate.

We applied a motor neuron differentiation protocol based on the addition of retinoic acid and sonic hedgehog to both wild-type mouse embryonic stem (ES) cells and to mouse ES cells that express mutant human SOD1. This allowed us to produce large numbers of the two corresponding types of motor neurons. We then developed a high content (automated microscope) screen designed to identify factors that prolong survival of either wild-type or mutant motor neurons, where motor neuron apoptosis was initiated by withdrawal of neurotrophic factors from the culture medium. Bioactive small molecule collections containing several thousand individual compounds were employed in these screens. Hits were identified and clustered separately by either suspected mechanism of action or by chemical structure (Fig 1). Screening of each motor neurons type (ie. Wild-type or mutan SOD1) rave rise to a unique list of hit compounds, and we are very interested in the compounds that are able to promote the survival of both types of motor neurons (overlapping hits). We are currently pursuing compounds or cellular targets/pathways of particular mechanistic interest that promote both mouse and human motor neuron survival. Further work will be needed to determine the characteristics of compounds most likely to be of therapeutic value.

Huntington’s Disease

Huntington’s Disease (HD) is an autosomal dominant neurodegenerative disease, ultimately leading to chorea, behavioral and emotional changes, and dementia. HD is one of the more common genetic disorders, affecting approximately one American in 10,000. It is characterized by degeneration of medium spiny neurons (MSNs) in the striatum. Finding effective treatments for HD will be greatly accelerated by having access to large numbers of human MSNs expressing mutant Huntingtin. As a first step, we are trying to develop a protocol to direct the differentiation of mouse embryonic stem (ES) cells into MSNs using high content screens of small molecule and morphogen libraries. We anticipate that a similar protocol, based on the one we are establishing, can be applied with minor modifications to human ES cells, thereby providing disease-relevant cells to use in HD therapeutic screens.

Transcription factors known to be expressed sequentially by the developing striatum in vivo, beginning with Sox1 and FoxG1, are being utilized to mark the progressive differentiation of ES cells. To begin, we screened Sox1+ neural progenitors and identified agents with two types of activity. Several compounds increased the expression of Sox1 in the progenitors, without changing their number. We are currently testing the differentiation capacity of these cells. We have also tested approximately 2000 compounds in three separate sessions for their ability to enhance expression of the forebrain progenitor marker FoxG1, either by increasing the number of cells that express nuclear FoxG1 or by increasing the relative intensity of nuclear FoxG1. We also used an additional telencephalic marker, the transcription factor Six3, as a readout for about 320 of the compounds. All compounds were tested in three concentrations for the first step. From this data, we have identified 160 compounds from the original 2000 screened that we will further test in a 10-point dose curve. We chose compounds by weighting them as more promising if they could increase nuclear signal in both markers tested. Additionally, we grouped potential hits by similarities in the chemical composition of the compounds and weighted them as more promising if similar chemistries hit in multiple rounds of testing.


Viral transduction of three transcription factors can reprogram adult somatic cells to induced pluripotent cells that possess many of the known characteristics of embryonic stem cells. Unfortunately, the potential for aberrant integration of foreign genetic material into the host’s genome and the expression of reprogramming factors with oncogenic potential limit the therapeutic utility of these cells. Such limitations could be overcome by identifying small molecules that compensate for the removal of each of the transcription factors by either inducing their endogenous expression or by mimicking their activities during the reprogramming process. To this end, we designed a series of high-content screens, each based on the activation of Oct4::GFP in mouse embryonic fibroblasts. Using carefully chosen annotated compound libraries, we identified small molecules capable of selectively replacing Sox2, Klf4 and Oct4. Induced pluripotent cells derived using the identified small molecules appropriately express all stem cell markers, can be directed to differentiate in vitro, form teratomas composed of tissue from all three embryonic germ layers, contribute to the inner cell mass of a blastocyst, and yield live chimeric offspring. Thus, somatic cells reprogrammed using the identified small molecule replacement compounds are likely bona fide induced pluripotent cells.

The small molecules that replaced Sox2 and Klf4 could be used to reprogram adult somatic cells with the efficiency achieved through overexpression of the transcription factors themselves. In addition, the use of such mechanistically annotated compounds allowed us to identify novel pathways in the reprogramming process. Finally, the chemical replacement of critical reprogramming factors demonstrates the feasibility of reprogramming adult somatic cells to a pluripotent state using solely small molecules. We are currently testing combinations of the small molecule replacers to provide a complete chemical reprogramming method.


SCRB 190

Understanding Aging: Degeneration, Regeneration, and the Scientific Search for the Fountain of Youth with Amy Wagers

This lecture and discussion course will explore the fundamental molecular and cellular mechanisms that govern organismal aging and contemporary strategies to delay or reverse this process.