Cocaine exposure impairs multineage hematopoiesis of human hematopoietic progenitor cells mediated by the sigma-1 receptor

Prenatal exposure to cocaine is a significant source of fetal and neonatal developmental defects. While cocaine associated neurological and cardiac pathologies are well-documented, it is apparent that cocaine use has far more diverse physiological effects. It is known that in some cell types, the sigma-1 receptor mediates many of cocaine's cellular effects. Here we present a novel and concise investigation into the mechanism that underlies cocaine associated hematopoietic pathology. Indeed, this is the first examination of the effects of cocaine on hematopoiesis. We show that cocaine impairs multilineage hematopoiesis from human progenitors from multiple donors and tissue types. We go on to present the first demonstration of the expression of the sigma-1 receptor in human CD34 + human hematopoietic stem/progenitor cells. Furthermore, we demonstrate that these cocaine-induced hematopoietic defects can be reversed through sigma-1 receptor blockade.

depressed. Interestingly, we found that S1R blockade relieved the impairment and allowed for normal hematopoiesis to proceed. Thus, our data suggest that stimulant exposure can impact human hematopoiesis and that the S1R may play a role in this developmental process, beyond substance abuse.

Results
Nontoxic levels of cocaine exposure depresses multilineage hematopoiesis from human CD34 1 HSPC. We first sought to determine if exposure to physiologically relevant concentrations of cocaine is toxic to human hematopoietic stem progenitor cells (HSPCs). Cocaine treatment levels and exposure time were used based on previous studies that determined on the basis of physiological effect on T cells without any toxicity risks 22,26 . CD34 1 progenitors from fetal liver (FL) were incubated at varying timepoints with cocaine. Following treatment, a portion of the cells from both cocaine exposed and untreated cultures were examined by flow cytometry for cell death. Representative flow cytometry plots indicate no apparent differences in cell viability comparing untreated and cocaine treated cells at 24 and 48 hours post-exposure ( Fig. 1a, (Fig. 1a,b). There were no evident differences in the viability of cultures that were exposed to cocaine compared to those that were not. These data indicate that cocaine is nontoxic to HSPCs in culture at a concentration of 10 28 M. Since cocaine had no apparent impact on cell viability, we then addresed whether cocaine exposure alters the hematopoietic potential of HSPCs derived from either FL, cord blood (CB), or adult bone marrow (BM). A portion of the cells from cocaine treated and untreated cultures were plated in complete methylcellulose and analyzed two weeks later for the development of hematopoietic colonies. In plates derived from HSPCs treated with cocaine, total colony number was severely reduced compared to untreated HSPCs (Fig. 2a,2b,2c). In FL derived cultures from three separate donors, cocaine treated cells gave rise to an average of 23 1/2 3 total colonies while untreated cells gave rise to an average of 87 1/2 17 total colonies. This difference is statistically significant by two tailed t-test with a P value ,0.001. In CB derived cultures from three separate donors, cocaine treated cells gave rise to an average of 25 1/2 3 total colonies while untreated cells gave rise to an average of 93 1/2 17 total cells. This difference is statistically significant by two-tailed ttest with a p value ,0.001. In BM derived cultures from three individual donors, cocaine treated progenitors gave rise to an average of 13 1/2 3 total colonies while untreated progenitors gave rise to an average of 44 1/2 5 total colonies. This difference is statistically significant by two-tailed t-test with a p value ,0.001.  Table 1 shows that in almost every case derived from FL and CB, each type of colony from each donor, from both source tissue types, HSPCs that have been exposed to nonlethal levels of cocaine are impaired for development of all cell types. There is no apparent pattern of donor/colony types in which development is not impaired in a statistically significant fashion following drug exposure. All erythroid colonies from all but donor 3 showed that difference between the number of colonies that develop from untreated compared to cocaine treated progenitors. It is difficult to interpret the colony numbers of both the erythroid and the granulocyte/macrophage mixed cultures due to the fact that the numbers of these colony types in untreated cultures were quite low. However, there is strong significance in the differences in total colony number from this donor. BM derived cultures show similar results. While there were no differences in erythroid colonies that developed from cocaine treated compared to untreated progenitors, the development of almost all other myeloid colonies from all donors were suppressed in cultures derived from treated progenitors. Overall, these data indicate that cocaine exposure suppresses myeloid development from HSPCs derived from multiple tissues and types.
The sigma-1 receptor is expressed in human CD341 HSPCs. As previous work has implicated the sigma-1 receptor in mediating the cellular effects of cocaine 22,26 , we sought to determine if this receptor is present in human HSPCs. RT-PCR anaylsis for the sigma-1 receptor transcript in FL derived HSPCs from four separate donors indicate that it is indeed expressed in blood progenitors (Fig. 3a). The average of three RT-PCR reactions from donor 1 indicated 0.14 1/2 0.0078 copies of S1R per copy of GAPDH, while it was 0.17 1/2 0.0083 for donor 2, 0.11 1/2 0.013 for donor 3, and 0.14 1/2 0.0088 for donor 4. S1R blockade alleviates cocaine-induced suppression of hematopoiesis. To elucidate the mechanism by which cocaine suppresses hematopoiesis, we expanded our cocaine exposure experiments to include blockade of the sigma-1 receptor. FL derived HSPCs from three separate donors were either left untreated or treated with cocaine, S1R blockade, or both for 48 hours and then plated in Table 1 | Hematopoietic colonies derived from untreated or cocaine exposed CD34 1 cells from either fetal liver, cord blood, or adult bone marrow. Donors 1-3 represent fetal liver, donors 4-6 represent cord blood, and donors 7-9 represent adult bone marrow. complete methylcellulose. Two weeks later, plates were scored for hematopoietic colony development. The total number of colonies that developed in cultures derived from cells that were treated with either S1R blockade or cocaine and S1R blockade were indiscernible while cultures that were treated with cocaine alone show significantly reduced total colony numbers (Fig. 3b). In this set of experiments, untreated HSPCs from all three donors gave rise to an average of 56 1/2 13 colonies per plate while those exposed to cocaine gave rise to only 12 1/2 2 colonies per plate. HSPCs that were treated with only the receptor block gave rise to 51 1/2 12 colonies and HSPCs that were treated with receptor block and cocaine gave rise to 51 1/2 13 colonies per plate. There were no significant differences in the total colony number that developed from untreated, receptor block treated, or receptor block and cocaine treated HSPCs. Phenotypic analysis indicates that neither treatment with S1R block or S1R block and cocaine appears to alter hematopoietic lineage commitment (Fig. 3d). To confirm that treatment with the receptor block is not toxic to developing hematopoietic colonies, individual colonies from each of the treatment conditions were examined by microscopy (Fig. 3c). Table 2 shows that only cocaine exposure results in a statistically significant loss of developmental potential. Interestingly, macrophages were the only lineage that was not impaired with cocaine treatment. Granulocyte, macrophage, and mixed granulocyte/macrophage colonies were clearly suppressed by cocaine exposure of HSPCs from all three donors. There were no statistically significant differences in the colony forming potential between untreated HSPC and those treated with just S1R blockade or both S1R blockade and cocaine. Together these data clearly show that the S1R can mediate changes in human hematopoiesis.

Discussion
In conclusion, our studies strongly demonstrate that cocaine exposure has far reaching effects on human development and is detrimental to early and intermediate stages of human hematopoiesis. We show that in HSPCs derived from both fetal liver and cord blood, cocaine exposure results in depressed hematopoietic potential as determined by colony forming assays. Moreover, we show that blockade of the sigma-1 receptor restores normal hematopoietic potential. Together these data establish the novel concept that cocaine exposure modulates human hematopoiesis.
It was previously reported that acute cocaine exposure induced terminal erythrocytosis as measured by hematocrit and reticulocyte count 18 . While this may seem to contradict the data presented here, it is important to note that erythroid colonies that develop in methylcellulose are not composed of terminally differentiated red blood cells but a mixture of intermediate-late erythroid progenitors. The investigators reported an increase in hematocrit but not reticulocyte count which would indicate an increase in terminally differentiated erythrocytes but not one in the production and/or release of immature red blood cells. Furthermore, the study consisted only of cocaine users who reported to the emergency department with drug associated acute chest pain. Therefore, we do not believe this represents a accurate picture of cocaine exposure.
Interestingly, macrophages seemed unaffected by cocaine exposure in the last set of experiments. This suggests the possibility that a macrophage precursor may be tolerant to the effects of cocaine exposure and demands further study. As this was not observed in the first set of experiments, more detailed analysis will be required. It is important to remember that the CD34 1 population is composed of diverse phenotypes and each sample is unique it its distribution of these. Taken together, these points clearly indicate that each subpopulation of HSPCs should be studied under the same conditions. This work opens the door to further study of the impact of drugs of abuse on human hematopoiesis. An early study indicated that mice deficient in the mu-opiod receptor, which mediates the cellular effects of a number of drugs of abuse, showed increased developmental potential and were in a much higher cycling state of HSPC relative to their wild-type counterparts 27 .
As this may potentially have downstream effects on the immune system of children exposed to cocaine in utero, S1R blockade may represent a potential protective treatment for women at high risk for drug abuse during pregnancy. Moreover, it is possible that some of the impaired development that results from in utero cocaine expose could be due to impaired blood development.

Methods
Ethics statement. Peripheral blood mononulear cells and cord blood were obtained at the University of California, Los Angeles in accordance with UCLA Institutional Review Board (IRB) approved protocols under written informed consent using an IRB-approved written consent form by the UCLA/CFAR Virology Laboratory and was distributed for this study without personal identifying information. Human fetal tissue was purchased from the UCLA/CFAR Gene Therapy Core, was obtained without identifying information and did not require IRB approval for use. All experiments and agents used were carried out and used in accordance with the UCLA Institutional Biosafety Committee (IBC).
Isolation and culture of human hematopoietic stem cells. Mononuclear cells were obtained from the buffy coats of either processed fetal liver or cord blood or adult bone marrow (Allcells, Alameda, CA) following Ficoll-Hypaqe density gradient separation. CD34 1 HSPCs were sorted from buffy coats by MACS using CD34 1 magnetic beads (Miltenyi Biotec, Auburn, CA, USA) and LS columns (Miltenyi Biotec) following manufacturer's protocol. Cells were cultured in StemSPAN TM serum-free media (Stem Cell Technologies) supplemented with stem cell factor (Life Technologies, Grand Island, NY, USA), thrombopoietin (Life Technologies), IGF- Real-time RT-PCR. RNA was isolated from cells using the RNeasy kit (Qiagen) following the manufacturer's protocol. PCR reactions were performed using the One Step RT-PCR Kit for Probes (Biorad, Hercules, CA, USA) and FAM labeled primerprobsets for SIR and GAPDH transcripts manufactured by Advanced Biosystems Inc. Twenty ml reactions were prepared for each sample with 5 ml of target RNA. All samples were run in triplicate. Samples were analyzed along with standards of known concentration to allow for quantification. All reactions were carried out on an IQCycler thermalcycler (Biorad). Cycling conditions were: 1 cycle of reverse transcription for 10 min at 50u C, one denaturation cycle for 3 min at 95uC, followed by 40 amplification cycles. Analysis was performed with iQ5 software (Biorad).
Flow cytometry. The Live/DeadH Aqua dye (Life Technologies) was used following manufacturers protocol to determine viability in HSPC cultures. Briefly, 1 3 10 6 cells were harvested from untreated and cocaine treated cultures at the indicated timepoints and washed twice and then resuspended in 1 mL of phosphate-buffered saline (PBS). One uL of dye was added to each sample, mixed, and then incubated in the dark for 30 minutes. Cells were then washed and fixed for 15 minutes in PBS-37% formaldehyde. Fixed cells were maintained in 500 uL PBS-37% formaldehyde at 4uC in the dark until flow cytometry analysis. All stained cells were analyzed together on an LSRFortessa TM (BD Biosciences, San Jose, CA, USA) flow cytometer.
Statistics. All statistical analysis was performed with Graphpad Prism software (La Jolla, CA, USA). P values , 0.05 were considered significant.  FL derived HSPCs from four separate donors. S1R expression is graphed relative to GAPDH expression and the average of three replicate reactions is plotted. (B) Methylcellulose colony forming assays of FL derived HSPCs that were either untreated, cocaine treated, S1R block treated, and cocaine 1 block treated. Each mark represents the score on a single plate, and the bar marks the mean. (C) Representative images of hematopoietic colonies that developed from S1R-block treated (top) or untreated (bottom) HSPC. (D) Phenotypic analysis of colonies that developed from HSPCs that were either untreated, cocaine treated, S1R block treated, and cocaine 1 block treated. Red boxes/E erythroid colonies, blue boxes/G granulocyte colonies, black boxes/M macrophage colonies, green boxes/GM mixed granolocyte/macrophage colonies. www.nature.com/scientificreports