Therapeutic Effects of Genetically Modied Wharton’s Jelly Mesenchymal Stem Cells Expressing Erythropoietin on Breast Cancer-related Anemia in Mice Model

Cancer-related anemia (CRA) negatively inuences cancer patients’ survival, disease progression, treatment ecacy, and quality of life (QOL). Current treatments such as iron therapy, red cell transfusion, and erythropoietin-stimulating agents (ESAs) may cause severe adverse effects including hemolytic transfusion reaction and the possibility of host immunity against rhEPO. Therefore, development of long-lasting and curative therapies is highly required. Combined cell and gene therapy platform can introduce a new route for permanent production of erythropoietin (EPO) in the body with various degrees of clinical benets and avoiding the need for repeat treatments. In this study, we developed cell and gene therapy strategy for in-vivo delivery of EPO cDNA via genetic engineering human Wharton’s jelly mesenchymal stem cells (hWJMSCs) to long-term produce and secret human EPO protein after transplantation into the mice model of CRA. To evaluate CRA's treatment in cancer-free and cancerous conditions, at rst, we designed recombinant breast cancer cell line 4T1 expressing herpes simplex virus type 1 thymidine kinase (HSV1-TK) by a lentiviral vector encoding HSV1-TK and injected into mice. After 3 weeks, all mice develop metastatic breast cancer associated with acute anemia. Then, we administrated ganciclovir (GCV) for 10 days in half of the mice to clear cancer cells. Meanwhile, we designed another lentiviral vector encoding EPO to transduce hWJMSCs. Following implantation of rhWJMSCs-EPO, the whole peripheral blood samples were collected from the tail vein once per week for 10 weeks which were immediately analyzed for the measurements of EPO, hemoglobin (Hb), and hematocrit (Hct) plasma levels. The blood analysis showed that plasma EPO, hemoglobin (Hb), and hematocrit (Hct) concentration signicantly increased and remained at a therapeutic level for >10 weeks in both treatment groups which indicates that the rhWJMSCs-EPO could improve CRA in both cancer-free and cancerous mice model.


Introduction
Cancer-related anemia (CRA), as a common consequence of tumor burden, occurs in more than 30% of cancer patients at diagnosis time which can reach 90%, when the patients go under aggressive chemoradiotherapy [1,2]. The mechanisms contributed to CRA are included in chemotherapy-induced anemia (CIA), blood loss, iron de ciency, erythropoietin de ciency due to renal disease, and bone marrow involvement with the tumor [3]. Due to CRA's negative effect on survival, disease progression, treatment e cacy, and the patient's quality of life (QOL), developing effective therapy for CRA is in great demand.
Current treatments for CRA include iron therapy, red cell transfusion, and erythropoietin-stimulating agents (ESAs) [4,5]. Erythropoietin (EPO) is a 30.4 kDa glycoprotein hormone primarily produced by the fetal liver and adult kidney, which plays an important role in the body such as erythropoiesis, tissueprotective effects, and immune regulatory effects on immune cells [6]. The administration of recombinant human erythropoietin (rhEPO) as a key regulator in the production of functional red blood cells is approved to treat the anemia of patients with long-lasting diseases such as cancer and renal failure with clinical bene ts in correcting hemoglobin levels and markedly reduced the required number of blood transfusion and so, patients bene t from advantages, such as improved cardiac function, enhanced exercise capacity, and better QOL [7]. However, rhEPO administration may cause some adverse effects, including the possibility of host immunity against rhEPO, frequent self-administration by the patients which may not know the correct injection method and high cost [8].
Gene therapy via designing the plasmid DNA and viral vectors encoding the EPO gene introduced an attractive research area for treating anemia. In preclinical studies, gene therapy for direct delivery of EPO gene into animal models' skeletal muscle has shown a signi cant increase in EPO and erythropoiesis; however, life-threatening polycythemia and host immune responses to viral vectors limit its utilization in the clinic [9,10]. Cell therapy is therapy in which viable cells such as primary, stem or progenitor cells or stem cell derivatives are injected or implanted into a patient [11]. Cell-based therapies were studied as a treatment option in animal and clinical phases since 1990 to address incurable diseases, such as autoimmune, skeletal, cardiovascular, neurological, ophthalmologic, and blood diseases which showed satisfactory safety and e cacy pro les [11][12][13]. Furthermore, the combined cell and gene therapy platform which we propose here is emerging as a potential alternative treatment to the traditional pharmacologic and also direct gene therapy; since it can permanently produce therapeutic proteins in the body with various degrees of clinical bene ts and avoiding the need for repeat treatments [14,15]. An effective cell and gene therapy protocol approach to deliver the EPO have more clinical and economic bene t than the repeated injection of EPO protein. Mesenchymal stem cells (MSCs) as desirable cell carriers can be easily obtained, expanded, and genetically engineered to express and secret therapeutic proteins in-vivo [14,16].
Human Wharton's jelly mesenchymal stem cells (hWJMSCs) are multipotent stem cells that showed the potential to differentiate into mesodermal, ectodermal, and endodermal lineages [17]. hWJMSCs have led to promising outcomes in preclinical and clinical studies due to their limited heterogeneity, ease of their isolation and culture, availability in several tissues, and ability to self-regenerate [18]. Also, compared to adult and fetal stem cells, hWJMSCs show a higher proliferation rate and minimum stimulation of immune and in ammatory systems [19]. So, they have newly emerged as an appropriate therapeutic vehicle for gene therapy and drug delivery. Their therapeutic applications in various disease models, including in ammatory and autoimmune diseases, and cancer are being studied [20][21][22]. However, there are still some challenges that need to be addressed for the successful application of hWJMSCs including age-related telomere shortening at higher passages, morphological changes and loss of their differentiation ability, and rapid death of the transplanted cells [23].
This investigation genetically modi ed hWJMSCs to long-term produce and secret human EPO protein after transplantation into CRA mice. We used breast cancer cell line 4T1 to develop mice model of CRA as it causes tumor-associated acute anemia [24]. To evaluate CRA's treatment in cancer-free and cancerous conditions, we genetically altered breast cancer cell line 4T1 to express HSV-TK and inject into the mice. After con rming anemia induced by breast cancer, we eliminate cancer cells by administrating ganciclovir (GCV) in half of the mice. Following implantation of rhWJMSCs-EPO, plasma EPO, hemoglobin (Hb), and hematocrit (Hct) concentration signi cantly increased which indicate that the EPO-transduced hWJMSCs could improve the anemia of cancer in both cancer-free and cancerous mice model and can provide supporting evidence for future studies as a valuable therapeutic tool for the treatment of anemia. (Gibco, USA), penicillin (100 units/ml)/streptomycin (100 mg/ml) (Invitrogen, Carlsbad, CA, USA). Finally, the cells were incubated in a humidi ed atmosphere with 5% CO 2 at 37°C.
2.2. Gene Design. The coding sequences (CDS) of EPO (Accession#: XM_001468996.1) were retrieved from GeneBank, NCBI. The sequences were synthesized by Genscript, USA, and were incorporated into pUC57 plasmid. In this experimental study, a dual promoter lentiviral vector, pCDH-513B, was purchased from System Bio, USA. The rst promoter is the cytomegalovirus (CMV) promoter with a downstream multiple cloning site (MCS) used for gene cloning. The second promoter is the EF1a promoter which regulates the expression of CopA-GFP (copepod green uorescent protein) (cGFP) and puromycin resistance genes.
2.3. hWJMSCs Isolation and Characterization. Human healthy umbilical cord (UCs) Wharton's jelly tissue (n= 1) was collected from full-term newborn in Vali-e-Asr Central Hospital and processed after obtaining the mother's informed consent and approved by the Ethics Committee of Fasa University of Medical Sciences (IR.FUMS.REC.1397.177). It was washed in phosphate-buffered saline (PBS) to eliminate the blood clots, disinfected, and cut into 1-2 mm lengths, and then were expanded on culture dishes with a minimum of Dulbecco's Modi ed Eagle Medium containing Nutrient Mixture F-12 (DMEM/F12) medium (Gibco, USA), supplemented with 30-40% FBS. Then, it was incubated at 37°C with 5% CO 2 which routinely monitored. After one week, solid umbilical cord pieces were removed, and cell migration was evaluated under the invert light microscope. Upon reaching 90% cell con uence, in the second changing medium, the adherent cells detached by the addition of 0.25% Trypsin-Ethylene Diamine and re-plated, usually at 4_6 days' intervals [25]. For adipogenic, osteogenic, and chondrogenic differentiation, a six-well plate was cultured with 1× 10 5 cells per well. In the third passage, after reaching a 40-45% cell con uence, an adipogenic differentiation medium (Gibco, USA), osteogenesis supplement (Gibco, USA), and chondrogenic supplement (Gibco, USA) were added to the basic medium, respectively. After about 20 days, lipid droplets were visualized using Oil Red O (Sigma-Aldrich, USA) staining.
Successful osteogenic differentiation was veri ed by Alizarin Red (Sigma-Aldrich, USA) staining. Safranin-O (Sigma-Aldrich, USA) staining was used to determine the presence of proteoglycans (PGs). In the control group, hWJMSCs were grown in the culture medium without adipogenic, osteogenesis, and chondrogenic supplement. Passage 3 WJMSCs were used in all experiments. A small number of undifferentiated hWJMSCs (passage 3, 10 5 cells) were analyzed using BD FACSCalibur ow cytometry (BD Bioscience, USA) to evaluate surface markers expressed by hWJMSCs. After adding speci c antibodies at the recommended concentrations, the tubes were incubated in the dark at room temperature for 30-60 minutes. Then, ow cytometry analysis was performed to study two positive markers _CD105 and CD90_ and two negative markers _ CD45 and CD34_ and data were analyzed using FlowJo (version 7.6.1) software.
Trono lab protocol, CaPO4 transfection of HEK293T cells was performed with some modi cations using the following amounts of DNA: 21 μg transfer/control vector, 10.5 μg pMD2.G vector, 15 μg pMDLg/pRRE vector, and 13 μg pRSV-Rev vector which all were dissolved in HEPES buffered water to reach 921μl.
Then, 33μl Tris-EDTA (TE) buffer was added, and the mixture was strongly mixed and kept at room temperature for 3 minutes. Then, 105 μl CaCl2 2.5 M was added, and the mixture was strongly vortexed and left for 3 min for making DNA-CaCl2 interaction. Then, 1050 μl HEPES 2X was added while the mixture was being vortexed. HEK293T cells (2×10 6 cells) medium was changed 2 h before transfection. Early-passage HEK293T cells (passages under 15) with 80 % con uency was co-transfected by plasmids in a way that 2100 μl transfection master mix was added per 10 cm HEK293T cells as a droplet to all areas of the plate. Then, they were incubated at 37 °C in 5 % CO 2 for 16 h.
After 24 hours, the transfection e ciency was assessed by GFP expression and visualized by an inverted uorescent microscope (Leica, German). We selected ve elds randomly under a uorescent microscope, and the percentage of GFP-positive cell numbers determined the transfection e ciency to total cell number. Mediums containing the virus were collected after 24, 48, and 72 h following transfection which were passed through 0.25μm pore lters to remove cellular debris. Recombinant lentivirus concentration was measured by the addition of polyethylene glycol (PEG) 600 50%, NaCl4 M, and PBS to recombinant viruses inside polypropylene bottles which stored at 4°C for 1.5 hours. Then, the tubes were centrifuged at 15000 rpm for 15 minutes at 4°C. To determine the titration of recombinant lentiviral, the number of GFP positive cells were counted using ow cytometry according to the equation "TU (Transduction Unite/ml) = [F × C/V] × D" in which F is the frequency of GFP-positive cells, C is the total number of cells in the well at the time of transduction, V is the volume of inoculum in mL, and D is the lentivirus dilution factor. Fresh recombinant titrated viruses at the volumes of 1000, 500, 100, 50, 20, and 0 μl were used for transducing hWJMSCs and 4T1 cells.

Transduction and Viability
Assay. The second passage of hWJMSCs and 4T1 cells were cultured at low con uency of 30-40% in a 6-well dish and incubated at 37 °C, 5 % CO 2 overnight. Then, they were transduced by rLV-GFP, rLV-EPO, and rLV-TK to generate rhWJMSCs, rhWJMSCs-EPO, and r4T1-TK respectively. Cell transduction was evaluated using a uorescent microscope, 72 hours after the transduction, and was compared with non-transduced MSCs and 4T1. Puromycin (1.5 μg/ml) selection started 72 hours after the transduction at passage 3 for the next 5 days. Then, we analyzed them morphologically by phase-contrast microscopy to con rm that transduction with lentiviral vectors has not changed the morphology of WJMSCs.
For MTT assay, we cultured 5×10 3 cells from transduced and normal hWJMCSs and 4T1 cells in 96-wells plates. After 1 day, we added MTT reagents and incubated them for 4 hours. Also, to measure r4T1-TK cells' sensitivity to GCV, we add 20 μg/ml GCV to the r4T1-TK cells culture medium. Then, we added the dimethyl sulfoxide (DMSO) to terminate the reaction and the plate was read at 570 nm wavelength using BioTek Instruments (Vermont, United States) microplate reader.
2.6. Western Blot Analysis. EPO and TK gene expression was con rmed after the transduction by western blotting assay. rhWJMSCs-EPO and r4T1-TK supernatants were collected 72 hours after the transduction. Evaluating the protein concentrations was examined using a BCA Protein Assay Kit (Thermo Fisher, USA). Equivalent amounts of proteins (30 µg/lane) were loaded onto 12% sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) and then transferred onto nitrocellulose membranes (Bio-Rad, USA). The membranes were blocked using 5% non-fat milk and immunoblotting was performed using antibodies against EPO and TK (Santa Cruz, USA). Proteins of interest were detected using HRPconjugated sheep anti-mouse IgG antibody (Abcam, ab6785). Finally, the protein bands were visualized using chemiluminescence (ECL) reagent, and the integrated optical density (IOD) of each protein band was measured. The internal standard β-actin adjusted IOD values. 2.7. CRA Mice Model. The mice were obtained from the laboratory animal center of Pasteur Institute of Iran. Male and female BALB/c mice (6-to 8 weeks old, n=60; weight, 18-20 g; n=10 mice/group) were housed and treated in a pathogen-free environment with the access to autoclaved food and water ad libitum according to national rules on animal experiments. Also, the mice were maintained on an iron-su cient diet for 2 weeks before the injection of tumor cells or saline, which continued until the end of the study[26]. 5 ×10 5 recombinant 4T1-TK cells diluted in 100 μl PBS were injected into the right fourth mammary gland of mice using a 25-gauge needle [27]. The tumor was palpably detected after one week of injection in all 60 mice; thereafter, twice a week routinely, the tumor volume was measured until the end of the study. Because r4T1-TK tumors were metastatic within 2-3 weeks after injection, to evaluate the consequences of tumor progression on erythropoiesis compare to the control group, peripheral blood samples were obtained from the tail vein at week 3 and were analyzed using a Sysmex XT-2000i automated hematology analyzer (Sysmex Corp., Hyogo, Japan) for the measurements of red blood cell (RBC) count, reticulocyte numbers, Hct, and Hb concentration to con rm anemia in mice. Those r4T1-TK-bearing BALB/c mice quali ed for developing the anemia were randomly divided into 6 groups (n= 10 mice/group). In this study, taking into account our published ndings in cancer-free and cancerous mice, three groups of animals were treated intraperitoneal (IP) injection with Ganciclovir (GCV, 100 mL) at 75 mg/kg twice daily for 10 continuous days 2.8. Implantation of rhWJMSCs-EPO. We evaluated our anemia treatment protocol in six groups of animals, all of which had anemia (3 cancer-free groups and 3 cancerous groups) according to the following instructions: A: cancer-free mice that received a moderate dose of rhWJMSCs-EPO to evaluate the effects of treatment during cancer suppression. B: control cancer-free mice that received control rhWJMSCs. C: control cancer-free mice that received PBS. D: control cancerous mice that received rhWJMSCs-EPO to evaluate treatment effects during the active phase of cancer. E: cancerous control mice which received rhWJMSCs. F: control cancerous mice which received PBS. Based on recent studies, 10×10 6 MSCs expressing EPO is considered a high dose of treatment which can develop polycythemia and 4.5×10 6 MSCs expressing EPO as a low dose of treatment has shown no correction of anemia [29]. So, to avoid polycythemia and also to achieve a better therapeutic response, we used a moderate dose of rhWJMSCs-EPO (~7×10 6 ) implanted into mice's skeletal muscle with a 29-G insulin syringe [29].
For laboratory measurements, the whole peripheral blood samples were collected from the tail vein once per week which were immediately analyzed for the measurements of Hct and Hb plasma levels. Moreover, the changes in plasma EPO levels in all CRA mice were detected by ELISA (Quantikine ELISA kit, R&D Systems, Minneapolis, MN, USA) according to the manufacturer's protocol.
2.9. Statistical Analysis. All data are present as mean ± standard error of the mean (S.E.M.). The statistical analysis of data was performed using the Student's t-test and analyzed by Prism software (GraphPad, San Diego, CA). P-value < 0.05 were considered statistically signi cant.

Results
3.1. Lentiviral Vectors Construction. Vectors containing EPO and HSV-TK were constructed using the DNA assembling method. Human EPO and HSV-TK cDNA were cloned in pCDH-513B-1 lentiviral vector ( Figure  1(a), (b)). Both genes were con rmed following enzymatic digestion and sequencing. The digestion of pCDH-EPO using XbaI and ApeI produced two bands of 1900 and 7800 base pair (bp) (Figure 1(c)), and digestion of the HSV-TK gene with SpeI and EcoRI showed two bands of 1100 and 8200 bp (Figure 1(d)). Sequencing was done for nal con rmation. In this construct, EPO and HSV-TK mRNA were transcripts from CMV promoter, and cGFP and Puromycin mRNA were transcriptions from the EF1a promoter.
3.2. hWJMSCs Isolation and Characterization. The hWJMSCs derived from human UC were isolated and cultured based on cell migration and surface attachment. In the third passage, the identity and properties of isolated hWJMSCs were veri ed by immunophenotyping of cell-surface antigens. According to ow cytometry analysis hWJMSCs were highly positive for CD105 and CD90 surface markers and negative for hematopoietic markers, CD45 and CD34 (Figure 2(a)). The broblastic-like morphology of the cells was con rmed under the inverted microscope. The hWJMSCs had a typical spindle shape with a capacity to differentiating into osteogenic, adipogenic, and chondrogenic lineages. The accumulation of lipid vacuoles was evaluated by Oil Red staining, Alizarin Red revealed calcium deposition, and the formation of PGs was con rmed via histological staining using Safranin-O (Figure 2(b)).
3.3. Lentivirus Production, In-vitro Transduction and Viability Assay. GFP reporter gene in the lentiviral vector was an index for transfection and transduction e ciency. Transfer and control vectors were cotransfected in early-passage HEK293T cells (passages under 15) with the helper packaging vectors based on CaPO 4 protocol. The transfection e ciency due to the GFP positive and negative cells counted under a uorescent microscope was higher than 90% (Figure 3(a)). Recombinant virus titrations were done as we described in materials and methods, and fresh rLV-GFP, rLV-EPO, and rLV-TK particle titration were approximately 2×10 6 , 1.8×10 6 , and 1.5×10 6 particles, respectively. We transduced 4T1 cells with rLV-TK and hWJMSCs with rLV-GFP and rLV-EPO and then cultured in a medium containing puromycin to separate them from non-transduced cells. The uorescent microscope visualized the expression rate of GFP. The transduction of hWJMSCs and 4T1 cells with concentrated and fresh recombinant viruses does not show any signi cant difference and we observed the broblast-like morphology by microscopy ( Figure 3(b)). Based on GFP positive and negative cells counted under uorescent microscope, transfection rate was higher than 90% and transduction rate was around 30% -40% (Figure 3(c)).
MTT assay showed that transduction and genome integration of transfer lentiviral vectors do not signi cantly affect the viability of both transduced hWJMSCs and 4T1 compared with non-transduced cells (Figure 3(d)). As TK activates by GCV, we tested the sensitivity of r4T1-TK cells to GCV by adding 20 μg/ml GCV to the r4T1-TK cells culture medium. Results indicated that r4T1-TK cells' survival decreased with the addition of GCV drug (Figure 3(d)).
Western blot results con rmed gene expression data at the protein level. Western blot analysis showed similar expression levels for β-actin protein in control and transduced hWJMSCs and 4T1 cells, compared to EPO and TK protein which was only overexpressed in transduced hWJMSCs and 4T1 cells, respectively (Figure 3(e)). These results showed that both proteins (EPO and TK) were transcribed and translated correctly.

CRA Mice
Model. Generating mice model of CRA, the recombinant r4T1-TK was injected into mice to develop breast cancer and anemia as a result of cancer development. At 14 days after injection, all tumor mice appeared in distress, as evidenced by lethargy and poor feeding secondary to tumor load. At rst, the consequence of tumor burden on erythropoiesis was evaluated via analyzing RBC counts, Hb concentrations, Hct, and reticulocyte counts in the peripheral blood samples of r4T1-TK tumor-bearing mice at week 3. r4T1-TK-bearing mice showed an anemic feature with lower RBC count, Hb, and Hct levels, as well as increased reticulocytosis, compared to control mice (Figure 4(a)). After con rming the presence of anemia in all tumor mice, we treated three groups with GCV drug, whereas the other three groups received PBS, as mentioned in Material and Methods. As expected, we observed a tumor suppression in mice that received a 75 mg/kg dose of GCV drug than the PBS groups in which tumor volume consistently increased (Figure 4(b)).

Effects of Treatment.
Our experimental procedures to evaluate the time course of the biological effect of EPO-secreting rhWJMSCs on the anemia of cancerous and cancer-free mice with experimental breast cancer is described in Materials and Methods which illustrated in Figure 4. Among cancer-free groups (A, B, and C), plasma EPO level in group A which received rhWJMSCs-EPO, signi cantly raised and reached number 105.4± 23 mU/ml at week 4, compared to its two control groups B and C ~ 15 ± 2.8 mU/ml which received rhWJMSCs and PBS, respectively ( Figure 5(a)). This model of change detected in cancerous groups (D, E, and F) in which plasma EPO level was 103.1 ± 21.6 mU/ml in group D, which received rhWJMSCs-EPO in comparison to its two control groups E and F ~17.5 ± 3.2 mU/ml which received rhWJMSCs and PBS, respectively ( Figure 5(a)). Although there was a signi cant difference in EPO level between each treatment group (A and D) and its control groups, we found a slight difference in EPO level between two treatment groups A and B.
After 7 days of intramuscular-implanted rhWJMSCs-EPO, an increase in Hb was achieved in A and D groups. However, compared to control groups that receive rhWJMSCs and PBS, they showed little difference in the rst week after the treatment (Figure 5(b)). The increase in hemoglobin becomes more substantial in two treatment groups per week 4 in which in group A, hemoglobin level rose to level 17.2 ±2.3 g/dl, to compare with group D 15.5 ± 1.8 g/dl. So, although hemoglobin level rose to its highest number in group D, it did not increase as much as the amount of hemoglobin in group A. Thus, we noticed a considerable difference in response to EPO treatment in cancer-free and cancerous groups. This different response to treatment was more noticeable after observing similar hemoglobin changes pattern among control groups (B and C vs. E and F) which didn't receive rhWJMSCs-EPO in which cancer-free groups B and C showed more increase in hemoglobin compare with cancerous groups E and F ( Figure   5(b)). It was present, while we had a similar pattern of changes in the rate of Hct increase in response to treatment. The Hct in group A, as in group D, reached the therapeutic level, more precise, to 60.3% ±2.7 in group A and 51.5% ±1.4 in group D, by approximately 4 weeks after rhWJMSCs-EPO transplantation, whereas the Hct was not signi cantly altered in control groups mice which received the control rhWJMSCs and PBS ( Figure 5(c)). Although the amount of Hct in cancer-free control groups (B and C) increased further compare to cancerous control groups (E and F), none of them reached therapeutic levels, similar to what was seen in Hb level of control groups ( Figure 5(c)).
In general, we found a signi cant correlation between the amount of plasma level of EPO and Hb and also Hct concentration. In both groups of A and D which received ~7×10 6 rhWJMSCs-EPO, the Hb and Hct which had declined from basal 16.1 ± 2.1 to 9.3 ± 0.7 and 51% ± 1.2 to 31.7% ± 0.5 respectively, approximately 3 weeks after r4T1-TK injection, reach the therapeutic level at week 4. As we measured the plasma level of EPO weekly, we found a decrease in its level after week 5, but stayed in therapeutic level for >10 weeks in groups A and D. As EPO decreased, the plasma level of Hb and Hct declined in both treatment groups. However, Hb and Hct concentration persist at a therapeutic level for >10 weeks for both treatment groups of A and D.

Discussion
Our study on CRA treatment has included cell and gene therapy approaches, whereby the EPO gene is transferred to hWJMSCs by lentiviral vectors in-vitro, and the cells are subsequently implanted in-vivo to serve as EPO-releasing vehicles to establish, if EPO signi cantly increased Hb and hct levels.
Several studies have reported using gene therapy to deliver the EPO gene using plasmid DNA and viral vectors such as adenoviral and AAV vectors [10,30,31]. By designing plasmid DNA expressing rat EPO gene and direct delivery into skeletal muscle of rat anemia model by subtotal nephrectomy, studies showed a noticeable increase in Hct [32,33]. The administration of adenoviral or AAV vectors into an animal model of anemia to deliver EPO resulted in an increase in plasma level of EPO and erythropoiesis which introduces viral vectors as a highly e cient gene delivery vehicle [34,35]; however, life-threatening polycythemia was reported [10,30]. The erythropoiesis response to therapy was proportional to the dose of plasmid DNA or viral vectors delivered. Moreover, host immune responses to these vectors and their transgene products are associated with potential health risks limiting their entry into the clinical phase [9,36].
One remedy to overcome the safety risks and the limitation of gene therapy approaches is using cells as delivery vehicles for plasma-soluble therapeutic proteins in-vivo like EPO, which allow us to quantify and control the serum level of EPO expressed by transduced cells through adjusting the number of implanted gene-modi ed cells secreting EPO to prevent severe polycythemia and also reduce the risk of systemic virus dissemination [37]. MSCs are promising candidates for gene delivery to treat hematological diseases like anemia, mostly due to their accessibility for genetic modi cation and the simplicity of their culture and expansion in vitro [16,38]. Indeed, some experimental studies were reported using the MSCs as a suitable delivery vehicle for therapeutic proteins in vivo [29,39]. Viral methods were widely used in the production of therapeutic protein by MSCs [40]. The main purpose of our investigation was to apply this biopharmaceutical approach for the EPO delivery in vivo for the treatment of CRA. In this study, we isolated MSCs from the human UCs as a good source of MSCs, because they can be harvested noninvasively in large numbers after birth with no ethical problems compared to MSCs derived from adults, have some advantages such as an improved proliferative capacity, life span, differentiation potential, and immunomodulatory properties which offer the best clinical utility [21,41].
We transduced hWJMSCs with an EPO-encoding lentiviral vector under highly controlled conditions in vitro to avoid any risk of viral dissemination in vivo. Transplantation of a moderate dose of rhWJMSCs-EPO (~7×10 6 ) into the CRA mice model's skeletal muscle resulted in a cell dose-dependent increase of EPO level that reached up to 100 mU/ml in both treatment groups (A and D) after 4 weeks. It remained high until the end of the study (>10 weeks) (Figure 5(a)). Both Hb and Hct increase in response to EPO in both groups A and D; however, the increase in Hb and Hct in cancer-free group A was more signi cant than the cancerous group D (Figure 5(b, c)). Also, cancer-free control groups (B and C), in comparison to cancerous control groups (E and F), which received control treatments (rhWJMSCs and PBS), showed a higher level of Hb and Hct. Whereas all control groups had a low level of EPO. Thus, we concluded that combined cell and gene therapy strategies for correcting CRA could be more effective if the cancer is treated at the same time. It is currently believed that chemoradiotherapy is the key means for treating cancer patients, making CRA worse, and other serious side effects [2,42]. So, developing advanced therapeutic procedures which precisely target cancer cells are in great demand. In this study, we engineered recombinant 4T1 cells expressing HSV-TK to inject and develop breast cancer-associated anemia in mice, followed by injecting GCV to clear almost all cancer cells expressing TK in three groups of anemic animals. As a result, showed (Figure 4(b)), three groups that receive GCV displayed a signi cant tumor regression compared to cancerous groups. So, we could evaluate the e cacy of rhWJMSCs-EPO in cancer-free and cancerous groups to correct CRA in which the cancer-free groups had no serious side effects or other organ damage due to cancer treatment by GCV. Although this cancer treatment has no clinical utilization and we just design it in our study to evaluate the effect of rhWJMSCs-EPO on CRA in the condition in which cancer is treated via a precise targeted-therapy method without serious side effects which we see in other methods like chemo-radiotherapy, this hypothesis has important clinical implications, because developing therapeutic methods that only target cancer cells and clear all of them without in uencing other tissues or organs, similar to what we did as an animal study, not only can improve CRA over time, but also can pave the other CRA treatment such as cell and gene therapy that we used in this study.
We observed a gradual decrease in plasma concentration of Hb and Hct during ~ 10 weeks in correlation to a decrease of EPO. It could be because, according to some studies, MSCs do not persist in the recipient organism for the prolonged periods [41,43]. According to some studies, the survival time of MSCs transplanted to the skeletal muscle varies from 72 hours to 8 months [44]. We hypothesized a second dose of rhWJMSCs-EPO transplantation could be associated with satisfactory therapeutic results that need further investigation in MSC engineering and therapy. Consequently, we will able to schedule treatment plans in which MSCs transplantation courses will be done with determined doses depending on the disease stage. Therefore, the cell and gene therapy approach used here in our study has its limitations as a long-term approach to CRA therapy.
Although cell and gene therapy approach to correct anemia of cancer is in its infancy and has its own limitations, this strategy for the sustained production and delivery of EPO using ex vivo gene therapy to genetically engineer hWJMSCs to produce EPO can eliminate many of the adverse effects and complications of current therapies for CRA (Table 1). Supraphysiologic response leading to polycythemia may develop after the rst transplantation of EPO-secreting hWJMSCs which may require resection of cells or, conversely as we mentioned earlier the modi ed MSCs lose their effectiveness over time and therefore, re-implantation may be required to enhance their clinical usage. as cell vehicle has some limitations such as the non-sustained release of the desired secretory protein due to inactivation of vector sequence following transplantation, and also, depending on the donor's age the expansion capability of normal broblasts may be restricted because they ultimately reach a stage when the cell division cycle slow down leading to cell aging which limits their clinical applications [47]. In contrast, hWJMSCs used in this study are attractive candidates due to their potential expansion ability, an immuno-privileged status, and easy access for collection, which afford us high-e ciency lentiviral engineering cells, culture, and utilization in vivo of selected modi ed cells[48-50].

Conclusion
Our data con rmed that administration of a moderate dose of rhWJMSCs carrying the EPO cDNA resulted in the elevated circulating level of EPO in the CRA mice model, which caused a persistent increase in Hb and Hct. So, a combination of hWJMSCs and lentiviral vector could be suggested as a novel cell and gene therapy approach for CRA. When this anemia treatment protocol is combined with a precise targeted-therapy for cancer cells, it will give the best results. Obviously, there are challenges to the clinical use of cell and gene therapies for diseases such as anemia. We will watch closely for clinical safety and e cacy using these methods for disease treatment that is needed to keep this eld ahead. The data used and/or analyzed to support the ndings of this study is available from the corresponding author on reasonable request.

Con icts of Interest
The authors have declared that no competing interest exists.  Isolation, expansion, and characterization of human Wharton's jelly mesenchymal stem cells (hWJMSCs). (a) Flow cytometry analysis of hWJMSCs showed a positive marker for CD105 and CD90 and negative marker for CD45 and CD34. Results showed that more than 95% of hWJMSCs are positive for MSCs markers and negative for hematopoietic stem cell markers. (b) Cultured hWJMSCs showed broblasts morphology, stained mineral calcium indicated osteogenic differentiation, oil droplets con rmed adipogenic differentiation, and Safranin-O staining con rmed the formation of PGs. Scale bar = 100 μm. Cancer-related anemia (CRA) in mice following injection of r4T1-TK and Effect of GCV on the tumor regression. (a) r4T1-TK-bearing mice showed an anemic feature with lower RBC count, Hb, and Hct levels, and increased reticulocytosis, compared to control mice (*p <0.05). (b) Three tumor-bearing mice groups were treated by 10 days of GCV, and the other three groups received saline. Tumor volumes were measured two times a week. Signi cant tumor regression was observed in three groups that received GCV. Data are presented as mean±SD. All tests were done in triplicate, *P < 0.05