Murine studies in models of colonic inflammation have demonstrated that a small percentage (1 – 5%) of mesenchymal stem cells (MSC) injected intravenously or intraperitoneally home to the inflamed colon1,2. This study shows that ultrasound-guided intracardiac injections of MSCs result in increased localization to the intestine.
Crohn's disease (CD) is a common chronic inflammatory disease of the small and large intestines. Murine and human mesenchymal stem cells (MSCs) have immunosuppressive potential and have been shown to suppress inflammation in mouse models of intestinal inflammation, even though the route of administration can limit their homing and effectiveness 1,3,4,5. Local application of MSCs to colonic injury models has shown greater efficacy at ameliorating inflammation in the colon. However, there is paucity of data on techniques to enhance the localization of human bone marrow-derived MSCs (hMSCs) to the small intestine, the site of inflammation in the SAMP-1/YitFc (SAMP) model of experimental Crohn's disease. This work describes a novel technique for the ultrasound-guided intracardiac injection of hMSCs in SAMP mice, a well-characterized spontaneous model of chronic intestinal inflammation. Sex- and age-matched, inflammation-free AKR/J (AKR) mice were used as controls. To analyze the biodistribution and the localization, hMSCs were transduced with a lentivirus containing a triple reporter. The triple reporter consisted of firefly luciferase (fl), for bioluminescent imaging; monomeric red fluorescent protein (mrfp), for cell sorting; and truncated herpes simplex virus thymidine kinase (ttk), for positron emission tomography (PET) imaging. The results of this study show that 24 h after the intracardiac administration, hMSCs localize in the small intestine of SAMP mice as opposed to inflammation-free AKR mice. This novel, ultrasound-guided injection of hMSCs in the left ventricle of SAMP mice ensures a high success rate of cell delivery, allowing for the rapid recovery of mice with minimal morbidity and mortality. This technique could be a useful method for the enhanced localization of MSCs in other models of small-intestinal inflammation, such as TNFΔRE6. Future studies will determine if the increased localization of hMSCs by intra-arterial delivery can lead to increased therapeutic efficacy.
Crohn's disease (CD) is a common chronic inflammatory disease of the small and large intestines and is thought to result from an inappropriate response of the host immune system to intestinal microbes7,8. Recent studies have shown that both murine and human mesenchymal stem cells (MSCs) can suppress inflammation in mouse models of intestinal inflammation1,3,4,5. There are multiple ongoing clinical trials that use human MSCs derived from bone marrow or adipose tissue to treat patients with inflammatory bowel disease (IBD), which includes CD9. Two routes for MSC therapy have been used in these clinical trials: one involves the systemic infusion (i.e., intravenously) of MSCs for luminal IBD (including CD), and the other involves the localized application/injection of stem cells in the fistula tract of patients with perianal CD. In a recent meta-analysis of MSC therapy for IBD, systemic (i.e., intravenous) MSC therapy for luminal IBD (including CD) was efficacious in up to 40% (95% CI: 7 – 79%) of patients, whereas the efficacy was much higher, at 61% (95% CI: 36 – 85%), when the MSCs were injected locally into the diseased CD fistula9. A recent phase III multicenter randomized placebo controlled trial of allogeneic adipose stem cells injected directly into the perianal fistula of CD patients showed statistically significant clinical and radiological evidence of perianal fistulae healing, corroborating the findings of the meta-analysis10. The reasons for the low efficacy of MSC therapy given intravenously for luminal CD has been inadequately investigated, but one reason may be the inadequate homing of MSCs to the site of inflammation. Murine studies in models of colonic inflammation have demonstrated that only a small percentage of MSCs (1 – 5%) injected intravenously reach the inflamed colon; the remaining MSCs are filtered by the lungs (first-pass effect)1,2,5,11,12. Multiple murine research studies have therefore used the intraperitoneal route (i.p.) for MSC administration in animal models of colitis4. However, they have also demonstrated that only a small fraction of cells reach and engraft the colon and that the efficacy is related to the secretion of soluble paracrine factors, like tumor necrosis factor-inducible gene 6 protein (TSG-6)2. The MSC mechanism of immunosuppression and healing involves a multipronged approach that involves paracrine; cell proximity-independent factors, like TSG-6; and cell proximity-dependent factors, like programmed death-ligand 1 (PD-L1); or Jagged 1. Therefore, MSC localization to the site of inflammation may result in increased efficacy9,13. In fact, a recent study showed that MSCs directly implanted at the site of colonic injury resulted in healing by secreting angiogenesis-promoting vascular endothelial growth factor (VEGF). On the other hand, a minimal healing effect was noted after intravenous injection5. To increase their localization to the site of inflammation (i.e., the small intestine in SAMP mice), this ultrasound-guided intracardiac injection technique for MSC administration in the left ventricle was developed. Image-guided injection ensures an accurate injection, which leads to a higher rate of success and to decreased morbidity and mortality rates. Moreover, the injection of MSCs into the left ventricle delivers them to arterial circulation, where they can reach the inflamed small intestine before becoming trapped in the lungs.
In this study, human bone marrow-derived MSCs (hMSCs) were used for injection in the SAMP-1/YitFc (SAMP) murine model of CD14. SAMP is a well-characterized spontaneous murine model of chronic inflammation that develops small-intestinal inflammation with nearly 100% penetrance14. The inflammation develops in response to microflora in the absence of any chemical, immunological, or genetic manipulation and closely resembles human CD11. Sex- and age-matched inflammation-free AKR/J (AKR) mice, the parental control mice of SAMP, were used in this study.
The hMSCs were isolated and expanded in the laboratory from bone marrow (BM) samples obtained from normal, unidentified donors after informed consent using validated and previously published protocols15,16. After isolation and expansion, the MSC ability in osteogenic, adipogenic, and chondrogenic differentiation was evaluated in the laboratory by multiple assays15. The osteogenic functional assay was performed by implanting ceramic cubes of hydroxyapatite/tricalcium phosphate matrix containing hMSCs subcutaneously in immunocompromised CB17-Prkdc SCID mice17. The cube assay demonstrates osteogenesis and chondrogenesis potential and is considered the ultimate test for evaluating individual MSC preparations17. To visualize hMSCs in vivo after injection, lentivirus was used to transduce hMSCs with triple reporter gene construct that consists of firefly luciferase (fl), monomeric red fluorescent protein (mRFP), and herpes simplex viral thymidine kinase (ttk), driven by a modified myeloproliferative sarcoma virus (mnd) promoter18. The firefly luciferase in the triple reporter is an enzyme that coverts injected luciferin to oxyluciferin in hMSCs and produces photons/white light. This is detected by the sensitive charge-coupled device (CCD) camera (bioluminescence) in an in vivo optical imaging system, enabling the visualization of live hMSCs in mice. Bioluminescence imaging (BLI) is a sensitive technique that can be used serially for tracking cells and for ex vivo analysis. The use of a strong mnd promoter drives the continuous expression of the triple fusion reporter gene construct and allows for the imaging of injected hMSCs for more than 16 weeks19. The hMSCs are difficult to transduce and have a low transduction efficiency. Using an optimized protocol, the hMSCs transduction efficiency was improved and the transgene expression was enhanced18. Using mRFP expression (one of the triple reporter genes) on flow cytometry, the ability to transduce hMSCs with a high efficiency of up to 83% was demonstrated. Differentiation assays and in vivo cube assays demonstrated the ability of the transduced hMSCs to differentiate into chondrocytes, adipocytes, and osteocytes17.
All mice experiments and procedures in the study were approved by Case Western Reserve University's Institutional Animal Care and Use Committee. The procedures were conducted in Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) accredited facility. The BM was aspirated under a University Hospitals Institutional Review Board-approved protocol at the stem cell core facility at Case Western Reserve University after informed consent.
1. Culture and Transduction of Human Bone Marrow-derived Mesenchymal Stem Cells (hMSCs)
NOTE: The isolation of hMSCs is described in detail in previous publications15,20.
2. Ultrasound-guided Intracardiac Injection of hMSCs into the Left Ventricle
NOTE: Use SAMP1/YitFc with established small-intestinal inflammation and age- and gender-matched AKR/J mice for the experiments. Maintain the SAMP and AKR mice under specific pathogen-free conditions, feed them standard laboratory chow, and keep them on 12-h light/dark cycles. Here, the ultrasound-guided cardiac injections were performed in a dedicated pathogen-free cardiac ultrasound room located in the CWRU animal research facility that was sanitized with a disinfectant and rinsed with 70% alcohol. Laboratory research personnel must wear personal protective equipment, including gowns, face masks with eye shields, and sterile gloves during the intracardiac injections. The body heat of the mouse must be maintained for the duration of the procedure.
3. Bioluminescence (BLI) Imaging of hMSCs
Figure 1 shows that hMSCs can be transduced with the triple reporter at a high efficiency, preserving their stem cell properties. Transduced hMSCs can be visualized in real time by BLI (Figure 1C, Figure 3). This novel, ultrasound-guided injection of hMSCs into the left ventricle of SAMP mice ensures a very high success rate of injection, allowing for the rapid recovery of the mice with minimal morbidity and mortality (Figure 2). The average time for injecting one mouse with hMSCs is approximately 10 min, and less than 8 – 9% morbidity and mortality were observed with this technique. The major reason for mortality is stroke and cardiac tamponade from hemorrhagic pericardial effusion. The ex vivo analysis performed 24 h after the intracardiac (intra-arterial) administration of hMSC confirms that this route of administration results in the homing of hMSCs to the inflamed small intestine of SAMP mice, as opposed to inflammation-free control mice (Figure 3C and D). Furthermore, hMSCs tend not to stay in the small intestine of inflammation-free AKR mice and start accumulating or getting trapped in the lungs (Figure 3B).
Figure 1. hMSCs can be Transduced with a High Efficiency for In Vivo Visualization and With the Retention of Stem Cell Properties. (A) Microscopic image of transduced hMSCs. Successful transduction of human MSCs, as determined by mRFP expression assessed with flow cytometry (B) and BLI (C). Transduced hMSCs retain stem cell properties, as confirmed by differentiation assays that demonstrate the ability of transduced hMSCs to differentiate into chondrocytes, adipocytes, and osteoblasts. For a functional assay, the ceramic cubes of hydroxyapatite/tricalcium phosphate matrix were implanted subcutaneously in immunocompromised CB17-Prkdc SCID mice to demonstrate the ability of hMSCs to form ectopic donor bone (D). Scale bars in A = 500 µm. Scale bars in C indicate units of BLI (103 photons/s/cm2/steradian). Scale bars in D = 200 µm. Please click here to view a larger version of this figure.
Figure 2: Instruments for Ultrasound-guided Injection. (A) A panoramic view of the instruments: cardiac ultrasound imaging system (red square), transducer mounted in its apposite holder, syringe holder and surgical table with nosecone for anesthesia (green square), and inhalation anesthesia machine with induction chamber (yellow square). (B) Ultrasonographic image of the syringe needle: the needle tip (red arrow) is clearly visible inside the left ventricle (yellow arrows). (C) Fresh arterial blood outflow into the syringe confirms the successful insertion of the syringe needle into the left ventricle prior to hMSC administration. Please click here to view a larger version of this figure.
Figure 3. hMSCs Localize to the Small Intestine after Intracardiac/Intra-arterial Administration in SAMP Mice. Representative images of 3 – 4 mice/group, euthanized 24 h after injection and from three independent experiments are shown. In vivo bioluminescence imaging demonstrates the presence of hMSCs in SAMP and AKR mice 24 h after injection (A). Ex vivo analysis in SAMP mice demonstrates that hMSCs preferentially localize to the small intestine (i.e., the site of inflammation) (D), whereas they tend to not stay in the small intestine of inflammation-free control AKR mice (C). In fact, after the first arterial passage, hMSCs start accumulating more in the lungs of AKR mice, as opposed to SAMP mice (B). A fraction of hMSCs can also be detected in the spleen, liver and lungs of SAMP and AKR mice (B,C,D). The scale bars indicate units of BLI (103 photons/s/cm2/steradian). Please click here to view a larger version of this figure.
This study describes a novel technique for the ultrasound-guided intracardiac injection of hMSCs in a small-intestinal mouse model of experimental CD. This technique has very high rate of survival and success, as the trajectory of the needle can be adjusted based on real-time, high-resolution images of the mouse left ventricle, provided by the ultrasound. The advantage of delivery to the left ventricle is that hMSCs are then distributed intra-arterially and bypass venous circulation, thus avoiding the aggregation of cells in the lungs. Previously, carotid artery catheterization for intra-arterial delivery was used to inject MSCs16,21. The carotid catheterization method involves surgery to expose the carotid artery in order to place a catheter inside the artery. The catheter is advanced towards the aortic arch for the transfusion of cells. The catheter is subsequently removed and the carotid artery ligated. This novel technique for the ultrasound-guided injection of MSCs in the left ventricle, as compared to carotid artery catheterization, is less invasive, is much faster (approximately 10 min/mouse versus 60 min/mouse), and has less morbidity and mortality (<10% versus 30%). The clinical translation of the mouse ultrasound-guided intracardiac injection of hMSCs to human clinical trials would be through the catheterization of the superior mesenteric artery (currently performed for the treatment of chronic mesenteric ischemia and intractable gastrointestinal bleeding) to deliver hMSCs to the diseased small intestine.
There are several considerations regarding modifications and troubleshooting of the ultrasound-guided intracardiac injection of hMSCs in the left ventricle of mice. A critical aspect of the procedure is mastering the use of the ultrasound to guide the injection needle into the left ventricle. Like any procedure, proper training, practice, and expert feedback are important components to successfully perform this procedure. Poor technique can lead to cardiac tamponade from hemorrhagic pericardial effusion and to death. To minimize cell clumping and the risk of embolization to brain (i.e., stroke), it is essential to resuspend the hMSCs prior to each injection. Before implementing this procedure, it is important to consult the veterinarian in charge of the local animal research facility to receive expert advice. Ensure that guidelines for the safe and effective use of animals are followed to minimize inadvertent injury to the mice.
To examine the biodistribution of hMSCs after the injection in mice, hMSCs were successfully transduced with triple reporter, with high efficiency. Differentiation assays performed demonstrated the ability of transduced hMSCs to differentiate into chondrocytes, adipocytes, and osteocytes18. For the osteogenic functional assay, the ceramic cubes of hydroxyapatite/tricalcium phosphate matrix were implanted subcutaneously in immunocompromised SCID mice and demonstrated the ability of hMSCs to form ectopic donor bone17,18. These differentiation and functional assays demonstrate that transduced hMSCs retain their stem cell properties. The results of this study show that transduced hMSCs used for intracardiac injection into SAMP mice localize to the small intestine, which is the site of inflammation in the SAMP mice. In the intestinal inflammation-free AKR mice, only a minority of hMSCs localized to the intestines 24 h after injection, implying that, in the absence of inflammation, hMSCs do not remain in the intestine. This data is in concordance with multiple other studies that have shown the ability of MSCs to migrate and engraft at the site of injury, such as the heart in acute myocardial infarction22 and the injured BM in a radiation-injured leg16. Like in other diseases, the local application of MSCs to colonic injury models has shown efficacy at ameliorating inflammation23. In a 2,4,6-trinitrobonzene sulfonic acid (TNBS) model of colitis in rats, Hayashi et al. injected BM-derived MSCs into the submucosa of the rat colon, surrounding the area exposed to TNBS, and demonstrated accelerated healing of the colitis23.
A recent study by Manieri et al. used a novel technique of endoscope-assisted injection of colonic MSCs in the distal colon, the site of colonic injury, in prostaglandin-deficient mice and showed their effectiveness in preventing the formation of penetrating ulcers5. SAMP is a model of small-intestinal inflammation, and these mice do not develop spontaneous inflammation in their colon. Our laboratory has developed and validated an endoscopic method for the assessment and quantification of colonic inflammation and tumor development in mice. However, endoscopic devices are not able to reach the small intestine in a live mouse24. The scoring system can only be used as terminal method for the assessment of inflammation in the SAMP mice after euthanasia. Therefore, the endoscopic method is not suitable for the local application of MSCs in the SAMP model. By using this novel technique of injection, enhanced localization of MSCs to the small intestine can be achieved. Whether the increased localization to the inflamed small intestine will translate into increased efficacy is not known and will be the focus of future studies. In addition to the SAMP model, the ultrasound-guided injection of MSCs would be a useful method for the enhanced localization of hMSCs in other models of small intestinal inflammation, such as the TNFΔRE6 murine model of experimental CD.
BLI is a sensitive and convenient technique that can be used serially to track hMSCs in vivo, but it does not allow for the precise quantification of hMSC homing. An advantage of using the triple reporter-transduced hMSCs in mice is that the enzyme ttk in the triple reporter allows for more quantitative positron emission tomography (PET) imaging19. Combining highly quantitative PET imaging (with a radiotracer probe) with a computed tomography (CT) scan (providing localization) allows for a quantitative estimate of the number of engrafted hMSCs at the diseased site, as the gamma counts are proportional to the number of the viable hMSCs labeled with the ttk gene. Exact quantification of hMSC localization can help to determine the percentage of hMSCs homing and therapeutic efficacy and will be the focus of future studies.
This technique, despite its many advantages, has some limitations: i) the requirement for an expensive cardiac ultrasound imaging system for mice, ii) the mortality and morbidity associated with intracardiac injection, and iii) the inability to be used for the slow infusion (i.e., over hours) of cells and other molecules. Despite these limitations, it has many benefits over the current technique of carotid artery catheterization for the intra-arterial injection of hMSCs, as highlighted above.
In conclusion, the results of this study demonstrate the effectiveness and convenience of the ultrasound-guided injection technique to improve the localization of hMSCs to the site of inflammation in the SAMP model of experimental CD. Future studies will determine if the increased homing of hMSCs by intra-arterial delivery can lead to increased therapeutic efficacy in models of small-intestinal inflammation.
The authors have nothing to disclose.
Maneesh Dave is supported by Department of Defense grant PR141774 and the Crohn’s and Colitis Foundation of America Career Development Award. The laboratory of Fabio Cominelli is supported by NIH grants DK042191 (F.C.), DK055812 (F.C.), DK091222 (F.C.), and DK07948 (F.C.). The laboratory of Arnold Caplan is supported in part by the L. David and E. Virginia Baldwin Foundation. The funding agencies had no role in the study analysis or writing of the manuscript. Its contents are solely the responsibility of the authors.
DMEM-LG | Gibco | 31600-091 | |
0.25% trypsin-EDTA | Gibco | 25-200-072 | |
Proteamine Sulfate | Sigma Aldrich | P4020-1G | |
D-Luciferin | Goldbio | luck-500 | |
Fetal Bobine Serum | Gibco | 26140-079 | |
Antibiotic-Antimycotic | Gibco | 15240-062 | |
PBS | HyClone | SH30256.01 | |
175 cm tissue cutlure flasks | Corning | 431080 | |
75 cm tissue cutlure flasks | Corning | 430720 | |
Centrifuge tubes | Crysalgen | 23-2265 | |
Tissue-Plus O.C.T. compound | Fisher HealthCare | 4585 | |
Cryomold Standard | Tissue-Tek | 4557 | |
Lenti-Pac HIV Expression Packaging System | GeneCopoeia | HPK-LvTR-20 | |
Povidone Ionidine swabs | Medline | MDS093901 | |
Hair removal cream | Nair | N/A | |
Isoflurane | Piramal Helathcare | NDC 66794 013 25 | |
Forceps | Fine Science Tools | 11200-33 | |
Dissection scissors (Wagner) | Fine Science Tools | Wagner 14068-12 | |
Puralube vet ointment | Puralube | NDC 17033-211-38 | |
Aquasonic 100 ultrasound transmission gel | Parker laboratories | 01 08 | |
Petri dish | Fisher Scientific | FB0875711A | |
SAMP mice | Cleveland Digestive Disease Research Core Centre | ||
AKR mice | Cleveland Digestive Disease Research Core Centre | ||
Vevo 770 imaging system | Visual Sonics, Toronto, Canada | ||
IVIS spectrum series system | PerkinElmer, Waltham, MA | ||
Living image software | CaliperLifeSciences, PerkinElmer, Waltham, MA | ||
Triple reporter | Kindly provided by Dr. Zheng- Hong Lee, CWRU, (citation 19) |