Blood Outgrowth Endothelial Cells overexpressing eNOS mitigate pulmonary hypertension in rats: A unique carrier cell enabling autologous cell-based gene therapy

Abstract

We have investigated a unique cell type, Blood Outgrowth Endothelial Cells (BOEC), as a cell-based gene therapy approach to pulmonary hypertension. BOEC are bona fide endothelial cells, obtained from peripheral blood, that can be expanded to vast numbers, and are amenable to both cryopreservation and genetic modification. We established primary cultures of rat BOEC and genetically altered them to over-express human eNOS plus green fluorescent protein (rBOEC/eNOS) or to express GFP only (rBOEC/GFP). We gave monocrotaline (MCT) to rats on day 0, and they developed severe pulmonary hypertension. As a Prevention model, we infused saline or rBOEC/GFP or rBOEC/eNOS on day 3, and then examined endpoints on day 24. The rBOEC/eNOS recipients developed elevated NOx (serum and lung) and less severe: elevation of RVSP, right ventricular hypertrophy, and pulmonary arteriolar muscularization and loss of alveolar density. As an Intervention model, we waited until day 21 to give the test infusions, and we examined endpoints on day 35. The rBOEC/eNOS recipients again developed elevated NOx and manifested the same improvements. Indeed, rBOEC/eNOS infusion not only prevented worsening of RVSP but also partially reversed established arteriolar muscularization. These data suggest that BOEC may be useful as a carrier cell for genetic strategies targeting pulmonary hypertension. Their properties render BOEC amenable to preclinical and scale-up studies, available for autologous therapies, and tolerant of modification and storage for potential future use in patients at risk for PAH, e.g., as defined by genetics or medical condition.

SUMMARY

Background:

Pulmonary hypertension (PHT) is a disease in need of improved therapeutics. Although there is interest in cell-based gene therapy for this disease, experimental results have been rather disappointing. We reasoned that Blood Outgrowth Endothelial Cells (BOEC) may be an advantageous cell type in which to overexpress eNOS and deliver it for therapeutic advantage.

Translational Significance:

In demonstrating that BOEC overexpressing eNOS exerted therapeutic efficacy for rat PHT in both prevention and intervention models, we have shown that this unique cell type (which has many advantages) should be considered as a promising carrier cell for gene therapy of PHT.

Brief Commentary

Background:

Pulmonary hypertension (PHT) is a disease in need of improved therapeutics. Although there is interest in cell-based gene therapy for this disease, experimental results have been rather disappointing. We reasoned that Blood Outgrowth Endothelial Cells (BOEC) may be an advantageous cell type in which to overexpress eNOS and deliver it for therapeutic advantage.

Translational Significance:

In demonstrating that BOEC overexpressing eNOS exerted therapeutic efficacy for rat PHT in both prevention and intervention models, we have shown that this unique cell type (which has many advantages) should be considered as a promising carrier cell for gene therapy of PHT.

INTRODUCTION

Pulmonary arterial hypertension (PAH) is characterized by increased pulmonary vascular resistance, right heart failure, and untimely death, with a median life expectancy of only 7 years after diagnosis [1]. Even with current therapeutic strategies, five-year survival is limited to approximately 65% [2]. Although PAH can emerge from disparate medical conditions, its complex development always involves vascular wall remodeling, inflammation, and pulmonary vasoconstriction [3]. Unfortunately, these complex tissue changes already have developed by the time PAH is diagnosed, so an optimal therapeutic would need to foster some reversal of established disease.

As a possible therapeutic target, consideration has been given to the endothelial dysfunction with deficient NO bioavailability that is a hallmark feature of PAH [4,5]. Benefits expected from improving local NO bioavailability include: improved vasodilation ability, anti-inflammatory effects, blunted platelet activation, and inhibition of smooth muscle cell growth [6]. Indeed, in humans PAH of various subtypes can respond somewhat to inhaled NO, but its short half-life renders this impractical for long-term therapy [7]. Therefore, there has been interest in trying cell infusion or modalities of eNOS over-expression in the hope of locally and durably increasing pulmonary vascular NO bioavailability, as a rational and biologically-based therapeutic approach to PAH [8–13].

To this end, several investigations have tried infusion of so-called “EPC” that are unmodified [14,15] or manipulated to overexpress a vasodilator, either eNOS [10] or adrenomedullin [16]. Although intended to mean “endothelial progenitor cell”, the acronym EPC has customarily been applied to cells generated using the method devised by Asahara et. [17]. However, it has been clearly demonstrated that the “EPC” generated from the Asahara method are, in fact, unrelated to endothelial cells [18–21]. Rather, they are partially transdifferentiated peripheral blood monocytes that have acquired some endothelial features but that retain myeloid features, e.g. strong positivity for CD14. Nonetheless, their study in experimental PAH has yielded promising results.

As an alternative, we here examine whether using Blood Outgrowth Endothelial Cells (BOEC, aka ECFC) [20] would appropriately target and deliver eNOS over-expression to exert salubrious effects. BOEC are obtained from peripheral blood using a very specific culture method [20,22]. BOEC are not progenitor cells, but they are the progeny of a circulating, marrow-derived, transplantable endothelial progenitor cell [22]. BOEC are bone fide endothelial cells, as documented by multi-parameter testing (described in Methods).

The present studies used rat BOEC (rBOEC) engineered to over-express human eNOS and infused into syngeneic rats given monocrotaline (MCT) for PAH induction. The results reveal significant efficacy for both prevention and intervention. The multiple features that render BOEC attractive as therapeutic gene carriers are noted in the Discussion.

METHODS

Methods are provided in greater detail in the online Supplemental Information file. All animal studies were approved by The Institutional Animal Care and Use Committee of the University of Minnesota.

Rat BOEC

Culture.

Using pooled blood from 1–2 week old Fisher F344 rat pups, we established cultures of rat Blood Outgrowth Endothelial Cells (rBOEC) using the same method we previously deployed to produce murine BOEC [23,24]. We confirmed endothelial identity by multiple parameters and markers.

Genetic modification.

From human umbilical vein endothelial cells we obtained a full-length human eNOS cDNA (3.8 Kb). Using a retroviral vector (modified MIRG) and serial cycles of transduction and FACS selection/enrichment, we engineered rat BOEC (rBOEC) to express GFP-plus-eNOS (“rBOEC/eNOS”) or GFP only (“rBOEC/GFP”). In characterization studies, each type comprised a single population of endothelial cells identified by cobblestone morphology, functional properties (uptake of acetylated LDL, tube formation in Matrigel, and production of NO), positive staining for endothelial markers and absence of staining for exclusionary markers (Fig. 1). In vitro, rBOEC/eNOS initiated greater tube formation in Matrigel (Fig. 1H vs. I) for which measurements are shown in Suppl. Fig. 1, and they produced greater NOx than rBOEC/GFP did (Fig. 1J).

Figure 1.

Characterization of rat BOEC (rBOEC). (A) Normal rBOEC, bright field microscopy. After transduction/selection, both rBOEC/GFP and rBOEC/eNOS exhibit: (B) GFP expression; (C) acLDL uptake; and expression of (D) VE Cadherin, (E) RECA-1, (F) Flk-1. By flow cytometry (G) rBOEC/eNOS comprised a single population positive for endothelial markers and negative for non-endothelial markers [open curves, isotype control Ig; shaded curves, test mAb]. (H, I) In Matrigel, tube formation after 24 hr was significantly more robust for rBOEC/eNOS (I) than for control rBOEC/GFP (H). (J) NOx release into culture supernatant after 20 hr was greater for rBOEC/eNOS (grey bars) than for control rBOEC/GFP (open bars) (_*, P<.05 for each), and the rBOEC/eNOS responded appropriately to addition of arginine (Arg) or L-NAME. Three separate trials of the above experiments were performed.

Study Protocols

These studies used 6-week-old Fischer 344 rats, both male and female, of admixed litters, with rats chosen randomly for assignment to study groups.

Monocrotaline (MCT) induction.

On day 0, we left some rats untreated and gave others MCT (75mg/kg, ip) to induce pulmonary hypertension [25]. For use within each study protocol the MCT rats were randomized to the three treatment arms.

Prevention Protocol.

On day +3 after MCT, we infused (via jugular vein) saline or control rBOEC/GFP or rBOEC/eNOS (cell dose of 1×10⁶ BOEC). Then, on day 24 we measured study endpoints.

Intervention Protocol.

After MCT induction on day 0, we waited until day 21 and then infused (via tail vein) saline or control rBOEC/GFP or rBOEC/eNOS (cell dose as above). Then, on day 35 we measured study endpoints.

Major study endpoints

We measured mean right ventricular systolic pressure (RVSP) using isoflurane anesthesia, intubation followed by ventilation with a rodent ventilator, and insertion of a 1.6F Pressure Catheter into the right ventricle via a limited thoracotomy [26]. We measured NOx in BOEC culture supernates, rat serum, and homogenized tissues using the amiNO-700 sensor. After euthanasia, we harvested organs and measured the RV/(LV+septum) weight ratio. For lung histology, we stained for smooth muscle actin to evaluate arterioles <10μm diameter, enabling their classification as non-muscular (NM), partially muscular (PM), or fully muscular (FM). Alveolar density, p-selectin staining and in vivo BOEC staining are detailed in the supplemental methods online. To quantify distribution of BOEC amongst different organs, we applied quantitative PCR for eNOS and/or GFP.

RESULTS

Phenotypic characterizations of the rBOEC, rBOEC/GFP, and rBOEC/eNOS all yielded similar results confirming their endothelial identity, shown for rBOEC/eNOS in Fig. 1A–G. Compared to control rBOEC/GFP, the rBOEC/eNOS exhibited more robust tube formation in Matrigel. (Fig. 1H, I; corresponding quantification is presented in Suppl. Fig. 1) and generated greater NOx (Fig. 1J). In separate studies, we tested the utility of BOEC in preventing PAH and for intervention in established PAH.

Prevention Model.

After receiving MCT on day 0 and the treatment infusions on day 3, rats were examined on day 24. Compared to untreated controls, rats given saline or control rBOEC/GFP showed little or no increase in NOx in serum or lung (Table 1A), and they developed: severe RVSP elevation (Fig. 2A), RV hypertrophy (Table 1A), increased P-selectin staining (Suppl. Fig. 3, panel L and M), abnormal pulmonary arteriolar muscularization (Fig. 2B), loss of alveolar number (Suppl. Fig. 4A), and failure to gain weight (Table 1A).

Table 1A:

Prevention model ^#

	Unmanipulated	Treatment Group
	Controls	Saline	rBOEC/GFP	rBOEC/eNOS	P∞
RVSP (mmHg)	21.9±1.6	49.8±6.5	48.3±10.4	35.6±4.3	0.003
RV/(LV+Septum)	0.26±0.03	0.47±0.06	0.45±0.10	0.38±0.07	0.054
Serum NOx (nmol/μl)	1.43±0.06	2.07±0.28	2.46±0.49	2.93±0.54	0.044
Lung NOx (nmol/mg)	477.5±9.9	420.0±50.4	367.3±132.4	660.7±203.8	0.001
Weight Gain (%)	69.3±8.2	57.2±12.2	51.4±21.1	67.6±16.3	0.036
Table 1B: Intervention modelϕ
	Unmanipulated	d+21	Treatment Group
	Controls	After MCT	Saline	rBOEC/GFP	rBOEC/eNOS	P∞
RVSP(mmHg)	23.2±2.6	67.5±17.2	93.8±9.0	92.0±16.9	58.3±6.5	0.01
RV/(LV+Septum)	0.26±0.01	0.40±0.04	0.64±0.07	0.63±0.06	0.61±0.09	0.313
Serum NOx (nmol/μ1)	2.77±0.63	1.81±1.26	4.26±0.97	4.79±1.19	7.51±0.96	0.003
Lung NOx (nmol/mg)	84.8±62.3	nd	119.7±67.6	99.2±22.6	141.7±42.5	0.048
Liver NOx (nmol/mg)	89.1±37.5	nd	117.5±19.8	171.6±25.4	196.9±29.1	0.095

^#Prevention model: MCT given on day 0; treatments given on day +3; endpoints measured on day +24.

^ϕIntervention Model: MCT given on day 0; treatments given on day +21; endpoints measured on day +35.

^∞P values compare rBOEC/eNOS with the rBOEC/GFP control.

Figure 2. Prevention model: Efficacy of early treatment with rBOEC/eNOS.

Some rats remained untreated (Cntrl; n=4). Others received MCT on day 0, and then on day 3 were given saline (n=10) or control rBOEC/GFP (n=9) or rBOEC/eNOS (n=10). When examined on day 24, the saline and rBOEC/GFP rats had severely elevated RVSP (A) and pulmonary arteriolar muscularization (B), while rats given rBOEC/eNOS developed far less severe changes. Muscularization scoring reports the % of arterioles that were non- (NM, open bars) or partially- (PM, grey bars) or fully- (FM, black bars) muscularized quantified from 9 random HPFs per animal, three animals per treatment group. To simplify, statistical comparisons are indicated only for the arterioles having no muscularization. (C-F) Histopathology of corresponding pulmonary arterioles stained for smooth muscle actin (brown) showed this to be (C) nearly absent in untreated rats, (D) very notable in saline rats, (E) improving slightly in control rBOEC/GFP rats, and (F) nearly normalized in rBOEC/eNOS rats; scale bar 50 μm. (G) Level of lung NOx; n=4,7,8,10 respectively. P values shown in graphs: *, P<.05; **, P≤.01; #, P≤.001.

In contrast, rats given rBOEC/eNOS developed increased serum and especially lung NOx (Table 1A). In parallel, they exhibited a lesser degree of RVSP elevation (Fig. 2A), RV hypertrophy (Table 1A), endothelial activation via P-selectin staining (Suppl. Fig. 3, panel N), and arteriolar muscularization (Fig. 2B). They also preserved alveolar density (Suppl. Fig. 4A) and gained weight appropriately (Table 1A).

Intervention Model.

After receiving MCT on day 0, rats were allowed to establish PAH. Then, on day 21 they were infused with the treatment options, and we subsequently examined them on day 35. The rats that had been given saline or rBOEC/GFP on day 21 exhibited significant worsening of RVSP (Fig. 3A) and RV hypertrophy (Table 1B) between days 21 and 35. Also, they continued to have arteriolar muscularization (Fig. 3B), increased P-selectin staining (Suppl. Fig 3, panel P and Q) and alveolar loss (Suppl. Fig. 4B).

Figure 3. Intervention model: Efficacy of rBOEC/eNOS for established PAH.

Some rats remained untreated (Cntrl) while others received MCT on day 0. Then on day 21 some were taken for testing, and the rest were given either saline or control rBOEC/GFP or rBOEC/eNOS. Animals were then examined on day 35. (A) It is evident that the already-severe RVSP on day 21 had significantly worsened by day 35 in the animals given saline or rBOEC/GFP. Strikingly, the rats given rBOEC/eNOS had not worsened at all between day 21 and day 35; and thus, were significantly protected from progression of RVSP; n=6,5,5,4,4 respectively. (B) Measures of arteriolar muscularization exhibited changes that paralleled the RVSP, quantified from 9 random HPFs per animal, three animals per treatment group. (Indeed, it appears that rBOEC/eNOS actually diminished arteriolar muscularization that had been there on day 21 (compare Fig 2B with Fig 3B). (C-F) Arterioles stained for smooth muscle actin (brown) showed this (C) nearly absent in untreated rats, (D) severe in saline rats, (E) not improved in control rBOEC/GFP rats, and (F) nearly normalized in rBOEC/eNOS rats. (G) Lung NOx values were less robust here than in the Prevention model, probably due to fewer animals. Yet rats given rBOEC/eNOS did have significantly higher NOx than those given rBOEC/GFP; n=6,4,5,4 respectively for both lung and liver NOx. P values shown in graphs: *, P<.05; **, P≤.01; #, P≤.001.

In striking contrast, the rats given rBOEC/eNOS on day 21 were fully protected from worsening of RSVP (Fig. 3A) and alveolar loss (Suppl. Fig. 4B), although there was no improvement in RV hypertrophy (Table 1B). Further, diminished p-selectin staining and endothelial activation was appreciated (Suppl. Fig 3, panel R). Remarkably, the treatment with rBOEC/eNOS caused partial regression of their established arteriolar muscularization, as is evident by comparing the 35 day treated value in Fig. 3B to the 21 day saline value in Fig. 2B. In parallel with these benefits, the rBOEC/eNOS rats had significantly higher NOx in serum and lung, and they trended higher in liver (Table 1B).

In these day 35 animals treated with either rBOEC/GFP or rBOEC/eNOS, a PCR-based search of major organs identified BOEC in liver, lung, spleen, bone marrow, kidney and heart (Suppl. Fig. 2B).

DISCUSSION

Pulmonary arterial hypertension may be a disease for which genetic manipulation offers a feasible approach. We here report successful mitigation of pulmonary arterial hypertension in the MCT rat model by infusing BOEC that over-express eNOS. This was true not only in our Prevention model but also with the Intervention model in which PAH had already become established before rBOEC/eNOS therapy. Indeed, in the Intervention model, rBOEC/eNOS actually induced some reversal of the already-established pulmonary arteriolar muscularization.

We studied BOEC because they have multiple properties that specifically suggest suitability for use in autologous cell-based therapies [20]. They are readily accessible from peripheral blood and are fully differentiated endothelial cells by multiple criteria (morphology, phenotype, functional testing, gene expression, organelle content). They expand robustly up to ~10¹⁹ cells in long-term culture [20,22]. They can be cultured using autologous donor serum, rather than xeno-proteins [20]; and they can be cultured using an entirely enclosed system [27]. Importantly, BOEC tolerate both genetic manipulation and cryopreservation. Thus, engineered BOEC could be prepared and stored for future anticipated use, as we demonstrated in a study of canine hemophilia [28].

In our experience with these cells, we successfully established BOEC cultures from human [22], dog [28], mouse [23], and rat (the present work). We also have done so from cow, sheep, pig, and rabbit (Lin and Hebbel, unpublished data). So BOEC-based therapeutics can be tested in pre-clinical and scale-up studies. Regarding durability, we observed continued transgene expression after 180 days in NOD/SCID mice given human engineered BOEC [29]; and we saw continued transgene expression after 425 days in dogs given engineered autologous BOEC [28] (these were the longest time points examined in the experiments). For genetic manipulation, BOEC have been successfully engineered using retroviral [28,29], Sleeping Beauty [30], and lenti-viral [31] vectors. Thus, their features suggest that modified BOEC could be prepared and stored for possible future use, e.g., for patients with substantial PAH risk identified by genetics or medical condition.

Of course, several caveats are relevant to this study.

We did not attempt to prove that the observed therapeutic benefit derived specifically from those eNOS-producing BOEC that took residence in the lung. Indeed, we already know that location of infused BOEC varies with time. In a study infusing autologous BOEC into dogs we found complete removal from the circulation within 5 minutes [28]. In a study infusing BOEC into mice, the great majority were initially entrapped in the lung, possibly mechanically since we were unable to implicate specific adhesion molecules [28]. However within 24 hours the BOEC were found widely dispersed in multiple organs, apparently having secondary trafficked [23,28]. This resulted in relatively few left in the lung. Yet, over the ensuing months the number of BOEC in the lung had again increased, as we know they do proliferate to some degree after engraftment [22].

For the present limited feasibility study we did not repeat these detailed tracking and localization studies. However, we did apply molecular testing at 14 days after BOEC infusion in the Intervention model and identified BOEC in multiple organs, including the lung. Interestingly, however, we found the greatest number BOEC residing in the liver. This perhaps raises the possibility that therapeutic effect in the lung could result from increased NO production in liver inducing a secondary effector that benefits the lung. Testing this hypothesis was beyond the scope of the present study.

The initial trapping of BOEC in the lung may raise concern about adverse effects on pulmonary flow and pressures. However, since engineered BOEC can be cryopreserved, a therapeutic approach could utilize multiple smaller doses.

Finally, we note that the beneficial effects observed in the present study were significant but only partial. However, we here made no attempt to optimize BOEC dose or schedule.

Thus, we suggest that BOEC engineered to over-expression eNOS should be further explored for possible deployment as an autologous, cell-based gene therapeutic approach to PAH.