A role for natural killer cells in the rapid death of... : Transplantation (original) (raw)
The massive and rapid death of isolated cultured donor myoblasts after injection in vivo remains a major hurdle for successful myoblast transfer therapy (MTT) designed as a clinical gene replacement therapy (1) for Duchenne’s muscular dystrophy (DMD). DMD is caused by defects in the gene encoding the subsarcolemmal protein dystrophin and is eventually lethal. After injection into host muscle, cultured donor myoblasts undergo rapid and massive cell death (2–4), in marked contrast with the excellent long-term survival (1 year) observed when donor myoblasts in (uncultured) sliced muscle grafts are implanted into host muscle (5,6). These data indicate that myoblast isolation and culture have adverse effects on the survival of donor myoblasts after transplantation into the in vivo environment. The extraordinary speed of the donor myoblast death is difficult to explain.
Recent quantitative studies show that initial survival of cultured donor myoblasts is enhanced three- to four-fold by antibody depletion of host CD4+/CD8+ or natural killer (NK)1.1+ cells before MTT (4). NK cells are prime candidates for such a lethal and rapid in vivo effect because they can bind and act on target cells within minutes (7). NK cells use cytolysis–cytokine production as major effector mechanisms (8), rapidly damage developing myotubes (9), and are associated with xenograft rejection (10). Antibodies used to deplete host lymphocyte function-associated antigen (LFA)-1, an adhesion molecule present on the surface of NK and other inflammatory cells, enhanced donor myoblast survival for up to 4 weeks (11), and antibody depletion of host NK cells also results in cardiac allograft acceptance in CD28-/- mice (12).
Because NK cells recognize target cells by way of a mechanism that is inhibited by expression of major histocompatibility complex (MHC) class I molecules (13), we hypothesized that altered MHC-I expression on donor myoblasts (because of tissue culture) provokes a host NK response. The murine cytomegalovirus (CMV) m144 protein has high homology to cellular MHC-I molecules (14) and confers resistance to cytolysis mediated by NK cells (14), and, therefore, manipulation of donor myoblasts to express m144 was designed to ablate the host NK response and enhance initial donor myoblast survival. These experiments complement ablation of the host NK cell response either by depleting antibodies (4) by the use of C57BL/6 (beige) host mice that have defective NK activity (often called “NK-deficient”) (15) and in mdx mice lacking perforin (16), a major mediator of NK activity (17). Our observations using antibody depletion of host immune cells (4) were extended to investigate the longer term (3 weeks) survival of donor myoblasts following initial specific depletion of NK1.1+ cells alone or in combination with CD4+ or CD8+, as well as sustaining the depletion regime for 3 weeks after MTT. Separate roles for CD8+ and CD4+ cells were investigated by antibody depletion of only CD8+ cells and in transgenic (GK5.5) CD4+-deficient mice (18).
A central role for host blood-borne factors (especially cells) in the death of donor myoblasts was critically tested by experiments using perfused and irradiated host mice. In all experiments, a Y-chromosome (male)-specific probe was used on total DNA extracted from injected muscles to quantify all male donor myoblasts in muscles of female host mice (4). These combined studies confirm a key role for the host NK cell response in the initial survival and long-term (3 week) maintenance of injected myoblasts and provide a strong foundation upon which to develop strategies to enhance MTT. These observations also have implications for survival of other cultured cell types after transplantation in vivo.
MATERIALS AND METHODS
Animals
Mdx, C57BL/10SnSc (normal parental strain for mdx), perforin deficient mdx, C57BL/Bg/Bg (beige, NK cell-deficient), and C57BL/6J (normal parental strain for C57BL/Bg/Bg) were obtained from the Animal Resource Centre, Western Australia. GK5.5 (anti-CD4 secreting) and BM-1 (normal parental strain for GK5.5) were a generous gift from Dr. Andrew Lew (The Walter and Eliza Hall Institute, Australia). All animal procedures were carried out in strict accordance with guidelines of the National Health and Medical Research Council of Australia and ethical approval from the University of Western Australia.
Primary Myoblast Cultures
Skeletal muscles were taken from the hind limbs of 4- to 6-week-old donor male C57BL/10SnSc, C57BL/6J, and BM-1 mice, and primary myoblast cultures were established by enzymic digestion as described previously (4).
Myoblast Injection
Myoblasts from primary cultures were adjusted to a concentration of 2.5×105 cells/10 μL in phosphate-buffered solution (PBS; pH7.2) on ice and injected into each tibialis anterior (TA) of 6- to 8-week-old female host mice as described previously (4). Cells from passages 3 to 5 were used routinely. C57BL/10SnSc donor myoblasts were used for injection into mdx and perforin-deficient mdx host mice, C57BL/6J donor myoblasts for beige host mice, and BM-1 donor myoblasts for GK5.5 host mice and samples taken from 0 hours (2–5 minutes) 3 weeks after injection. In other experiments, isolated TA muscles (5 minutes) from female mdx host mice were injected with either 2.5×105 (100%), 1.25×105 (50%), or 2.5×104 (10%) donor cells. This was done to circumvent the possibility of leakage of donor myoblasts and to compare the amount of male DNA recoverable from an injected muscle with that obtained from 2.5×105 myoblasts (designated 100%) pellet. These samples were then subjected to DNA quantification using the Y1 probe. DNA isolated from uninjected female host TA muscles was also mixed with varying amounts of donor male myoblast DNA (either 2.5×105 [100%], 1.25×105 [50%], or 2.5×104 [10%] cells) to investigate the possible masking effect of female DNA on slot blots.
Quantification of Donor Male Myoblast DNA
DNA extraction from whole TA muscles and quantification on slot blots using α32P deoxycytidine triphosphate (dCTP)-labeled Y1 probe is described in detail elsewhere (4). Donor male myoblast survival is represented as the percentage of male DNA recovered from samples when compared with that obtained from 2.5×105 cultured donor male myoblasts (designated as 100%). In all cases, identical donor myoblasts to those injected into host mice were used for 100% controls (n=5).
Myoblast Transfection with m144
C57BL/10SnSc male donor myoblasts (5×106) were transfected with a pCNDA-m144 construct encoding the full-length COOH-terminal c-myc-tagged MCMV m144 by way of electroporation (250 V, 960 μF, Bio-Rad Gene Pulser, Richmond, CA) as described previously (14). Two separate m144 transfection experiments and one control (peF plasmid only) were performed. Both pCNDA-m144 and peF control constructs contain a puromycin gene for growth in selective medium. After electroporation, myoblasts were resuspended in fetal calf serum (FCS)-supplemented medium and 48 hours later refed with selective medium (HAMS-F10 containing 20% FCS, 25 ng/mL puromycin, 12.5 ng/mL basic fibroblast growth factor [bFGF]). Cells suffered high (>95%) loss in selective medium after 72 hours, and successfully transfected cells took months to expand in culture. Viable clones were tested for m144 cell-surface expression by immunocytochemistry and Western analysis.
Western Analysis of m144
Untreated, peF control, and m144-transfected donor male myoblast whole-cell lysates were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotting. Blots were incubated with 15C6 murine anti-m144 monoclonal antibody (mAb) (14) at 1:200 dilution, followed by rabbit antimouse horseradish peroxidase (HRP)-conjugated secondary antibody (Dako, Carpenteria, CA) and detected using 4-chloro-1-naphthol chromogen.
Immunofluorescence
Test and control myoblasts were fixed with acetone and methanol (1:1), washed in PBS, and blocked using 10% FCS and 1% bovine serum albumin (BSA) in PBS for 30 minutes. Incubation with 15C6 mAb (14) at 1:200 dilution was followed by detection with antimouse immunoglobulin (Ig)G 6-([7-amino-4-methylcoumarin-3-acetyl] amino) hexanoic acid (AMCA)-conjugated secondary antibody (Jackson ImmunoResearch, West Grove, PA) at 1:400 dilution.
CD4+/CD8+/NK1.1+ Antibody Depletions
Host mice were injected intraperitoneally with 200 μL (450 μg) of mAb supernatant (YTS191 (anti-CD4+) and YTS169 (anti-CD8+) and PK136 (anti-NK1.1)), either alone or in combination, in serum-free medium as described (4) on days −5, −3, and −1, as well as a final injection 4 hours (day 0) before MTT. This was the initial depletion regime. For the sustained depletion regime, all three antibodies were also administered every second day for up to 3 weeks following MTT. Control depletion host mice were injected intraperitoneally with 450 μg of isotypic, nonspecific rat IgG (Sigma, St. Louis, MO) in serum-free medium.
Chromium-51 Release Assay for NK1.1+ Depletion
To confirm that mdx host mice had been depleted of NK1.1+ cells, a 51Chromium-release cytotoxicity assay was performed on splenocytes isolated from control mdx hosts using YAC-1 lymphoma target cells as described previously (4).
Fluorescence-Activated Cell Sorter Analysis of CD4+ and CD8+ Depletion
Spleens of depleted and control mice were subjected to fluorescence-activated cell sorter analysis (FACS), as described previously (4). CD4+ and CD8+ T-cell populations were analyzed using a 1:200 dilution of fluorescein isothiocyanate (FITC)-conjugated rat antimouse CD4+ (L3T4, Pharmingen, San Diego, CA) and rat antimouse CD8+ (53-6.7, Pharmingen) antibodies in PBS. Control cells from spleens and thymi of untreated mice were also tested.
Perfusion of Host Mice
To test whether host blood-borne factors were responsible for the death of the injected myoblasts, some host mice were perfused with saline before MTT. Female mdx mice were anesthetized by intraperitoneal injection of 100 μL of a mixture of ketamine–xylazine (100 mg, 10 mg/kg) and perfused thoroughly with PBS. TA muscles of perfused host mice were either left intact or surgically isolated (5 minutes), injected with 2.5×105 donor male myoblasts and 10 μL PBS, and subjected to DNA quantification using the Y1 probe.
Whole-Body Irradiation of Host Mice
To test whether circulating host leukocytes specifically were involved, some host mice received whole-body irradiation (WBI) 24 hours before MTT, which eliminates all leukocytes (19). Anesthetized mice were subjected to WBI or WBI followed by immediate bone-marrow reconstitution (BMR) 24 hours before MTT. Each mouse received 10 Gy (over 1000 Rads) of γ radiation, delivered at 2.5 Gy/min, to within an accuracy of ±2 mm using a Varian 600C Linear Accelerator (Palo Alto, CA) with a photon energy of 6 MV. For BMR, 6- to 8-week-old female donor mdx mice were killed and bone marrow extracted from the leg bones. Female mdx host mice were reconstituted within 1 hour of WBI with donor bone marrow cells (1×107/200 μL PBS) by way of tail injection.
RESULTS
Modification of Donor Myoblasts with m144
Expression of m144 protein in two clones of transfected C57BL/10SnSc myoblasts was confirmed using immunofluorescence (Fig. 1A), but no signal was detected in untreated or peF (plasmid only) control myoblasts. Western analysis of m144 in whole-cell lysates showed a 57 to 63 kD band in m144-transfected cells but not in untreated or peF control myoblasts (Fig. 1B).
Anti-NK molecule m144 expression enhances donor myoblast survival after myoblast transfer therapy (MTT). (A) Expression of m144 in 2 clones of transfected male C57BL/10SnSc donor myoblasts (immunofluorescence). Untreated or pEF (plasmid only) control myoblasts (no stain). Scale bar=20 μm. (B) Whole-cell lysates of donor myoblasts subjected to Western blot analysis using the anti-m144 antibody 15C6 showed the 57–63kD m144 protein (lane 5) expressed in both clones of m144-transfected cells (lanes 3 and 4) but absent in the untreated (lane 1) and peF (plasmid only) control myoblasts (lane 2). (C) Survival of two clones of male m144-transfected donor myoblasts enhanced immediately following injection into untreated female mdx host mice compared with untreated and peF (plasmid only) control male donor myoblasts, and numbers of m144 donor myoblasts are maintained for at least 3 weeks after MTT. Bars, standard deviation, 100% represents the total number of donor male myoblasts injected. *P <0.05 and **P <0.005, significant differences in values between treatments at each time point.
Survival of donor m144-transfected myoblasts (clone 1) was dramatically increased for up to 3 weeks in untreated mdx host mice compared with untreated or peF control transfected donor myoblasts (Fig. 1C). There was nearly a three-fold increase in the average survival of m144 transfected donor myoblasts at 0 hours (56% cf 16%), eight-fold at 1 hour, 22-fold at 24 hours and 1 week (168 hours), and still a 12-fold difference (29% cf 2.5%) at 3 weeks (504 hours) compared with untreated control myoblasts. Analysis of variance (ANOVA) showed a significant difference (P <0.05) between the numbers of donor myoblasts remaining in untreated mdx host mice at 0 hours, 1 hour and 24 hours, and (P <0.005) at 1 week and 3 weeks. Only at 1 and 24 hours was pEF control survival significantly (P <0.05) higher than untreated control. There was no significant difference between time points, indicating that increased survival of m144-transfected myoblasts was maintained for up to 3 weeks. To confirm these data, the experiment was repeated with a second clone of m144-transfected donor myoblasts. Again, initial survival was enhanced three-fold (46% cf 16%) at 0 hours and 19-fold (22% cf 2.5%) at 3 weeks (Fig. 1C).
Ablation of Host NK Response
Donor myoblast survival was dramatically increased for up to 3 weeks in mdx host mice after antibody depletion of host NK1.1+ cells (Fig. 2A). In host mice depleted of NK1.1+ cells before MTT (initial depletion only), there was about a three-fold increase in average survival of donor myoblasts at 0 hours (48% cf 16%). Numbers of donor myoblasts remained high compared with control (undepleted) hosts, with a 10-fold difference at 4 hours, 20-fold at 24 hours, 1 nine-fold at 1 week, and nearly a 10-fold difference (25% cf 2.6%) at 3 weeks. In sustained NK1.1+-depleted hosts, results were similar with nearly a four-fold increase (61% cf 16%) in the average donor myoblast survival at 0 hours, nine-fold at 1 hour, 22-fold at 24 hours, 21-fold at 1 week, and 18-fold (45% cf 2.6%) at 3 weeks, compared with control (undepleted) hosts. ANOVA showed that the differences between control (untreated) hosts and (initial and sustained) NK1.1+-depleted hosts were significant (P <0.05) at each time point. Numbers of donor male myoblasts in initial and sustained NK1.1+-depleted host groups were maintained over 3 weeks within each group and were not significantly different except at 3 weeks in which sustained depletion resulted in more donor myoblasts (P <0.05).
Effect of ablating the host natural-killer (NK) response on the survival of donor myoblasts following MTT. (A) Immediate donor myoblast survival is enhanced, and numbers are maintained for at least 3 weeks after MTT following injection into NK1.1+ antibody-depleted (single and sustained regime) female mdx host mice. However, absence of perforin did not assist in donor myoblast survival. (B) Initial donor myoblast survival is enhanced compared with C57BL/6J controls, but numbers gradually decline with time up to 3 weeks after MTT following injection into NK deficient C57BL/6J Bg/Bg (beige) female host mice. (C) Donor male myoblast numbers following MTT into untreated female mdx and female mdx host mice treated with nonspecific immunoglobulin (Ig)G isotype antibody follow the typical “death curve.” (D) Percentage recovery of NK cell activity following initial and during sustained depletion regimes. *P <0.05 and **P <0.005, significant differences in values between treatments at each time point; +P <0.05 and ++P <0.005, significant differences in values within each treatment group compared with time 0 hours.
To control for any possible nonspecific effects of antibodies on donor myoblast survival, mdx host mice were treated with a rat IgG isotype as a nonspecific control depletion regime. Numbers of injected donor male C57BL/10SnSc myoblasts after this treatment followed the typical “death curve” seen after MTT into untreated mdx hosts (4), in which initial donor myoblast survival was about 20% (of 100% originally injected) at 0 hours and gradually declined to about 5% after 1 week, as shown in Fig. 2C. There was no significant difference between untreated and nonspecific IgG-depleted host mice after MTT.
The efficiency of NK1.1+ cell depletion in mdx host mice was assessed using 51Chromium release assays and NK-sensitive YAC-1 target cells. Both NK1.1 antibody depletion regimes in mdx host mice (i.e., initial and sustained) almost totally depleted NK1.1+ cells at 0 hours: average 51Chromium release values ranged from only 5 to 14% (initial) or 5 to 12% (sustained) above spontaneous (background) control levels respectively. Between 1 to 3 weeks after the initial depletion, host NK activity had returned (with both initial and sustained depletion regimes) to 70 to 95% of that seen with control (undepleted) splenocytes (Fig. 2D).
In beige hosts, survival of normal C57BL/6J primary donor myoblasts was enhanced compared with injection into normal C57BL/6J hosts in which a characteristic death curve was seen over 3 weeks (Fig. 2B). In beige hosts (defective in NK cell activity), there was a two-fold increase in the average donor myoblast survival at 0 hours (42% cf 20%), and numbers of donor myoblasts were four-fold higher (21% cf 5.5%) at 3 weeks compared with control hosts. ANOVA showed a significant difference (P <0.05) in the numbers of donor myoblasts between beige and control hosts at all times. The decline in myoblast numbers within each treatment group (compared to time 0 hours) was also found to decline significantly (P <0.05) after 1 hour and further (P <0.005) from 1 to 3 weeks after MTT.
In perforin-deficient mdx mice, death of donor myoblasts was similar to that in control (undepleted) hosts at all times (examined up to 1 week), with average donor myoblast survival ranging between 5 to 10% (Fig. 2A). ANOVA showed no differences between perforin null mdx and control (untreated) mdx hosts.
Additional Role of Host CD4+ and CD8+ Cells
Analysis of CD4+ and CD8+ T cells from spleens by fluorescence activated cell sorting (FACS) showed effective depletion (10–20% of control untreated values) at the time of MTT. Previous studies show increases in donor myoblast survival from 1 hour to 3 weeks of 30 to 90% for CD4+ and 15 to 60% for CD8+ cells, respectively, after an initial antibody depletion regime (4). Following a sustained depletion regime after MTT, the numbers of CD4+/CD8+ cells (measured using anti-CD4+/CD8+ antibodies in combination) at 3 weeks was about 35% of that seen in nondepleted hosts. This indicates a significant recovery of CD4+ and CD8+ T cells over time, even with sustained depletion.
Increased survival of donor myoblasts was also seen immediately after injection into mdx hosts that had been pretreated with anti-CD4+ and anti-CD8+ antibodies (Fig. 3A). In host mice depleted of CD4+/CD8+ cells before MTT (initial depletion), there was nearly a three-fold greater survival of donor myoblasts at 0 hours (43% cf 16%). Despite numbers of remaining myoblasts gradually decreasing, at 3 weeks there was still nearly a four-fold difference (9% cf 2.5%) compared with control (undepleted) hosts. In sustained CD4+/CD8+-depleted hosts, similar numbers of donor myoblasts were present at 0 hours (39% cf 16%), and there was nearly a four-fold difference (9% cf 2.5%) at 3 weeks compared with control (undepleted) hosts. ANOVA showed that the increased numbers of myoblasts in CD4+/CD8+-depleted hosts were significant (P <0.05 compared with untreated controls) at each time point. There was no significant difference between initial and sustained CD4+/CD8+ host depletion at any time point, indicating no benefit in continually depleting the host mice. It is important to note that after initial or sustained CD4+/CD8+-depletion, numbers of donor myoblasts declined rapidly and significantly by 24 hours and at 1 week (P <0.05) and declined further by 3 weeks (P <0.005) after MTT. This indicates that, although CD4/CD8 antibody depletion did increase initial donor myoblast survival, this effect was very short lived, and protection was not maintained.
Effect of ablating the host CD4+/CD8+/NK1.1+ response on the survival of donor myoblasts following MTT. (A) Immediate donor male myoblast survival is enhanced, but numbers gradually decline over time following injection into CD4+/CD8+-depleted (single and sustained) female mdx host mice. Immediate donor myoblast survival enhanced, and numbers of donor myoblasts maintained at 3 weeks after MTT following injection into CD4+/CD8+/NK1.1+-depleted (sustained) female mdx host mice, but there is no additional benefit of CD4+/CD8+ depletion in combination with NK1.1+ depletion (B). Immediate donor myoblast survival is also enhanced following injection into CD8+-depleted (sustained) female mdx host mice, but there is no additional benefit of CD4+ depletion. (C) There is no difference in male BM-1 donor myoblast survival at any time point after injection into either the normal (parental) female BM1 or anti-CD4 secreting (GK5.5) host mice. Standard deviation is shown. *P <0.05 and **P <0.005, significant differences in values between treatments at each time point; +P <0.05 and ++P <0.005, significant differences in values within each treatment group compared with time 0 hours.
A sustained combination of anti-CD4+, anti-CD8+ and anti-NK1.1+ antibodies produced the same result as NK1.1+-depletion alone (Fig. 3A). In CD4+/CD8+/NK1.1+-depleted hosts, there were three-fold more donor myoblasts at 0 hours (51% cf 16%), 24-fold at 1 week, and a 16-fold difference (39.5% cf 2.5%) at 3 weeks compared with untreated control hosts. ANOVA showed that the differences between CD4+/CD8+/NK1.1+-depleted hosts and control (undepleted) hosts were significant (P <0.05) at each time point and that the numbers of donor male myoblasts in CD4+/CD8+/NK1.1+-depleted mdx hosts over 3 weeks was maintained. There were no significant differences in average donor myoblast survival between CD4+/CD8+/NK1.1+-depleted (sustained depletion) and (initial or sustained) NK1.1+-depleted hosts at each time point, indicating that there was no additive effect of combined host NK1.1+ and CD4+/CD8+-depletion.
To test the effects of depleting only CD8+ cells, dystrophic mdx hosts were pretreated with anti-CD8+ antibodies alone (Fig. 3B). In these CD8+-depleted hosts, there was a two-fold increase in donor myoblast survival at 0 hours (30% cf 16%); numbers of remaining myoblasts had declined rapidly by 1 week, but there was a three-fold difference (8% cf 2.5%) at 3 weeks compared with untreated control hosts (P <0.05). ANOVA showed that the differences between control (undepleted) hosts and CD8+-depleted hosts were significant (P <0.05) and that the remaining number of myoblasts was maintained until 1 week after MTT. Comparison with CD4+/CD8+-depleted hosts (Fig. 3B) showed no significant difference between CD8+ only or CD4+/CD8+ sustained depletion, indicating that CD4+ cells play little role in the initial survival of donor myoblasts.
To test the effect of removing CD4+ cells only, donor myoblasts were injected into transgenic GK5.5 host mice that are deficient in CD4+ cells and secrete anti-CD4 antibody in most tissues, including muscle (18) (Fig. 3C). The survival of normal BM-1 (the parental strain to GK5.5) primary donor myoblasts after injection into GK5.5 hosts was very similar to that seen in control BM-1 hosts, and both showed a characteristic death curve over 3 weeks (Fig. 3C). ANOVA analysis showed that there was no significant difference (P <0.05) in donor myoblast survival between GK5.5 and BM-1 host mice at any time point except at 1 week. These experiments support the idea that CD4+ cells play no role in the initial death of injected myoblasts.
Further Confirmation of Crucial Role for Circulating Host Factor
Female mdx host mice subjected to saline perfusion followed by MTT into intact or isolated TA muscles (Fig. 4A) showed that the vast majority (80%) of donor male myoblast DNA was recovered at 0 hours (also true for isolated muscles from untreated hosts). This is in marked contrast to the loss of male DNA usually seen in untreated host muscles (Figs. 1–4). These data show that replacement of the circulation with saline effectively removes some circulatory component that is involved with donor myoblast death following MTT. This further emphasizes that the normally low amounts of male DNA detected at time 0 hours cannot be accounted for by potential “leakage” of injected donor myoblasts or poor recovery of donor male DNA from muscle samples.
Testing a role for blood-borne host factors in the death of donor myoblasts. (A) Female mdx host mice were subjected to saline perfusion (or untreated) before MTT into intact or isolated tibialis anterior (TA) muscles. (B) Comparison of donor myoblast survival in whole-body irradiated (WBI) or WBI followed by bone marrow reconstituted (BMR) female mdx host mice. Standard deviation is shown. *P <0.05 and **P <0.005, significant differences in values between treatments at each time point; +P <0.05 and ++P <0.005, significant differences in values within each treatment group compared with time 0 hours.
To investigate the overall cellular nature of the host response to cultured donor myoblasts, female mdx host mice were subjected to (1) WBI or (2) WBI followed by BMR 24 hours before MTT (Fig. 4). Comparison is made with donor myoblast survival in the “standard” untreated control mdx host mice. With WBI, 80% of donor male myoblast DNA was present at 0 hours, demonstrating that WBI effectively “silences” the cellular immune response and provides almost complete protection for donor male myoblasts immediately after MTT. Reintroduction of the bone-marrow-derived leukocytes at 24 hours before MTT (by way of BMR) appeared to restore the adverse host immune response and resulted in the typical death of donor myoblasts following MTT, with approximately 25% remaining at 0 hours, 10% at 24 hours, and 16% at 72 hours, although ANOVA results showed that this drop was not significant.
DISCUSSION
A central role for some circulating host factor, especially bone-marrow-derived cells, in the rapid death of donor myoblasts is strongly supported by the MTT experiments in isolated, perfused, irradiated mice (in which donor cells survived) and BMR host mice (in which donor cells were killed). A central role for host NK cells in the rapid and massive death of cultured myoblasts was supported using two main strategies. Modification of donor myoblasts to express the m144 molecule provided dramatic protection from the initial death and maintained this protection for up to 3 weeks. It seems likely that the expression of m144 in donor myoblasts results in the ability of antigenic cells to mimic the expression of “self” antigens that result in the enhanced survival following injection. The striking protective effect of m144 expression in two separate clones of donor myoblasts lends support to the role of NK cells in the survival of donor myoblasts following MTT and to the hypothesis that altered MHC-I expression on donor myoblasts as a consequence of isolation and tissue culture provokes an adverse host response. This provides a strategy to directly treat donor myoblasts to circumvent the host NK cell attack as an alternative to manipulating the host’s NK immune response using NK1.1+ cell-depleting antibodies. The use of depleting antibodies to ablate the host NK cell response also produced initial and long-term protection of donor myoblasts. The fact that initial host NK cell depletion results in enhanced survival of donor myoblasts despite the reemergence of NK activity, strongly suggests that (at least in the mouse) a sustained depletion regime probably has no added benefit over an initial depletion before MTT.
The results indicate that host NK cells, rather than CD8 or CD4, are key players in this host immune response. The m144 treatment of donor myoblasts is particularly attractive because it seems to have wider applications and may provide a global protective effect against all NK cells. For example, similar protection was seen in another clone of m144 transfected donor myoblasts from transgenic (mixed lymphocyte culture [MLC]–insulin-like growth factor [IGF]-1) mice that overexpress IGF-1 (20) when injected into the parental FVB host strain (Hodgetts et al., unpublished observations, 2002). In contrast, depletion of the FVB host mice with anti-NK1.1 antibody was ineffective (Hodgetts, unpublished data) because only 60% of NK cells are NK1.1+ in this strain (21): such strain-specific differences emphasize the potential limitations of host NK-depletion strategies.
Further evidence to support a role for NK cells is provided by MTT studies in beige host mice that have impaired NK activity, in which donor myoblast survival was enhanced in comparison with untreated C57BL/6J control host mice. However, in this situation, donor myoblast numbers declined over time and were significantly reduced by 3 weeks. This decline in myoblast numbers might be accounted for by some compensatory mechanism that has developed in response to the lack of NK cell function in these mice (15).
Selective elimination of host perforin (a major mediator of NK activity) made no difference to the rapid death of donor myoblasts. This is readily explained by the compensatory use of alternative pathways (e.g., Fas or tumor necrosis factor [TNF]α) for NK activity (17). Recent studies show rejection of bone-marrow cells by NK cells even after perforin and Fas-mediated NK interactions were selectively eliminated (17). This complexity of in vivo cellular response necessitates the ablation of total NK activity for effective silencing of the NK response.
Beyond the initial survival of donor myoblasts (dramatically enhanced by these treatments), the numbers of myoblasts at later times may reflect a balance between cell survival, cell proliferation, and cell death. Available evidence indicates that myoblast proliferation is normally very low after MTT (2). This contrasts with marked replication seen in the highly mitogenic artificial environment of irradiated host muscles at 48 hours after MTT (2) and is confirmed by our similar observations of myoblast numbers in irradiated mice (Hodgetts, unpublished data, 2002). Regardless of the extent of myoblast proliferation, the long-term survival of donor muscle nuclei may depend on whether they fuse or become located beneath the basement membrane of myofibers because the interstitial environment appears unfavorable to myoblast survival. This proposal is based on studies with emigration of donor myoblasts from sliced muscle grafts that showed that over time (up to 12 weeks) more male nuclei (derived from muscle grafts) were observed located within myofibers compared with a scarcity of nuclei lying outside myofibers in the interstitial connective tissue in mdx muscles (5,6). Thus, the slight decline in numbers of donor nuclei seen at 3 weeks in hosts deficient in NK activity (antibody depleted or beige) may reflect the selective demise of myoblasts located outside myofibers (for reasons that might not necessarily involve immune cells).
NK cells secrete cytokines and chemokines that lyse target cells, which can be recognized within minutes (7). NK cells recognize target cells by way of a mechanism involving several surface receptors that are inhibited by combination with MHC-I receptors (13). Within minutes, MHC-I protein accumulates at the intercellular contact or “immune synapse” between NK cells and target cells and is organized into a ring around a central patch of the adhesion protein intercellular adhesion molecule (ICAM)-1 in association with LFA-1 (7). NK and other immune cells are LFA-1+, and depletion with anti-LFA-1 antibody enhances the survival of donor myoblasts following MTT (11,22). Therefore, altered MHC-I proteins on myoblasts resulting from exposure to tissue culture conditions [reviewed in (23)], or even during the cell harvesting process involving trypsin, may result in aberrant recognition by NK cell receptors, NK cell activation, and the destruction of donor myoblasts, mediated through cytokines and chemokines such as TNFα, interferon (IFN)γ, interleukins, perforin, and Fas (8,13,17,24), or alternatively through pathways independent of such molecules (17). Indeed, we have shown virtually absent MHC-I expression of donor myoblasts during culture by FACS (data not shown).
Ablation of the host NK response only resulted in protection of 60% of donor myoblasts. The death of 40% of injected donor myoblasts might be caused by the few residual NK1.1+ cells persisting after antibody depletion (up to 10%), yet incomplete protection is also seen in beige hosts that are deficient in NK activity. The demonstration that almost all of the injected donor myoblasts do survive initially when host mice are subjected to lethal WBI at 24 hours before MTT (Fig. 4D) strongly reenforces a role for circulating host-immune cells in donor myoblast death because such cells are eliminated within 24 hours of WBI (19). It seems likely that (in addition to NK cells) other immune cells, such as neutrophils, dendritic cells, macrophages, and mast cells, are also involved in the immune response directed against donor myoblasts at various stages following MTT [reviewed in (23)]. The problem of initial donor myoblast survival is probably multifactorial and complex. Dendritic cells are present in skeletal muscle (25), known to possess cytolytic activity and (in rats) even have markers shared by NK cells (26). It is possible, therefore, that NK-depletion indirectly affects such cells and their response to injected donor myoblasts. Similarly, there may be an indirect effect of CD4+/CD8+-depletion on NK cell function, which might influence the initial survival of donor male myoblasts following MTT. Indeed, there is evidence that some T cells (NKT cells) have similarities to NK cells because they share recognition systems that often involve receptors found on NK cells (27). More detailed analysis is required to address these complex interactions.
Additional depletion of CD4+/CD8+ cells did not further enhance the myoblast survival seen with NK1.1+-depletion alone. Although an improved (about 40%) initial survival of cultured donor male myoblasts was seen in mdx host mice subjected to initial or sustained CD4+/CD8+-depletion regimes, the numbers of donor myoblasts gradually decreased after time in both groups of CD4+/CD8+-depleted hosts. Even sustained depletion of host CD4+ and CD8+ cells was relatively inefficient long term because donor myoblast numbers declined by 3 weeks. The numbers of CD4+ and CD8+ T cells persisting because of the relative inefficiency of the antibody depletion (compared with NK) may alone account for the lower efficacy of these treatment regimes in myoblast survival.
However, it appears that CD8+ cells are important, whereas CD4+ cells seem to play little or no role in the initial death of donor myoblasts. This is demonstrated by depletion of host CD8+ cells alone and studies in CD4-deficient GK5.5 mice. It is unusual for CD4+/CD8+ cells to effect such a rapid immune response, although CD4+/CD8+ T cells are present in dystrophic (mdx) muscle (16) and have been attributed a role in the rejection of donor myoblasts in MTT (28). Our experiments do not exclude the possible participation of CD4+ or CD8+ cells in longer-term T-cell mediated responses that may be important in the later stage of MTT. The beneficial effect of CD4+/CD8+ or CD8+-depletion alone on donor myoblast survival cannot easily be explained by other than the possibility that some T cells may share certain properties of NK cells (27), which may be indirectly affected by the use of such antibodies, particularly CD8+-depletion alone.
The quantification of donor male myoblast DNA relies on comparison with 100% designated cells injected into each muscle. As reported previously (4), extreme care is taken to discard any samples that showed leakage of injected cells, and, although possible, this event is extremely rare. More importantly, the demonstration that almost complete recovery is obtained with isolated host muscles, perfusion of host mice, or after WBI endorses the suitability of using the number of injected cells as the designated 100% reference point. Indeed, such cell pellets were also used for 100% in the study in nude mdx host mice by Beachamp et al. (2). Other MTT studies show almost identical loss of donor myoblasts at 24 hours to that reported here (3,29), even when donor myoblasts are implanted in a fibrin clot (2), thereby eliminating the possibility of leakage.
The demonstration of a key role for host NK cells in the initial survival of transplanted cultured myoblasts marks a major advance in understanding (and hence overcoming) the savage and rapid host response to cultured transplanted myoblasts. It seems likely that activation of the host NK response may similarly be a major problem in the transplantation of other cultured cells for therapeutic purposes (e.g., islet cells, hepatocytes, and stem cells) (30). Modification of donor myoblasts (with molecules such as m144) or modulation of the host immune response (using NK-depletion regimes) to circumvent host NK attack provides a real opportunity to design strategies to enhance donor myoblast survival for future clinical trials of MTT.
Acknowledgments.
The authors thank Dr. Mariapia Degli-Esposti, School of Biomedical and Chemical Sciences, University of Western Australia (UWA), for her gifts of pCNDA-m144 and peF plasmids, the anti-m144 (15C6) monoclonal antibody, and her helpful discussions; Dr. Manfred Beilharz for his critical review of the manuscript; Mr. Daniel Andrews; and Dr. Matthew Wickstrom for his help and advice with the FACS analysis. The authors also thank Dr. Andrew Lew from WEHI in Melbourne, Australia, for his gift of the GK5.5 transgenic mice, and gratefully acknowledge the technical expertise and assistance of Marilyn Davies, School of Anatomy and Human Biology, UWA.
REFERENCES
1. Smythe GM, Hodgetts SI, Grounds MD. Problems and solutions in myoblast transfer therapy. J Cell Mol Med 2001; 5 ( 1): 33.
2. Beauchamp JR, Pagel CN, Partridge TA. A dual-marker system for quantitative studies of myoblast transplantation in the mouse. Transplantation 1997; 63 ( 12): 1794.
3. Beauchamp JR, Morgan JE, Pagel CN, et al. Dynamics of myoblast transplantation reveal a discrete minority of precursors with stem cell-like properties as the myogenic source. J Cell Biol 1999; 144: 1113.
4. Hodgetts SI, Beilharz MW, Scalzo T, et al. Why do cultured transplanted myoblasts die in vivo? DNA quantification shows enhanced survival of donor male myoblasts in host mice depleted of CD4+ and CD8+ or NK1.1+ cells. Cell Transplant 2000; 9 ( 4): 489.
5. Fan Y, Beilharz MW, Grounds MD. A potential alternative strategy for myoblast transfer therapy: the use of sliced muscle grafts. Cell Transplant 1996; 5 ( 3): 421.
6. Fan Y, Grounds MD, Garlepp MJ, et al. Increased survival, movement and fusion of myoblasts from sliced muscle grafts into skeletal muscles of T-cell depleted and tolerised dystrophic host mice. Basic Appl Myol 1997; 7: 231.
7. Davis DM, Chiu I, Fassett M, et al. The human natural killer cell immune synapse. Proc Natl Acad Sci U S A 1999; 96 ( 26): 15062.
8. Biron CA, Nguyen KB, Pien GC, et al. Natural killer cells in antiviral defense: function and regulation by innate cytokines. Ann Rev Immunol 1999; 17: 189.
9. Hohlfeld R, Engel AG. Lysis of myotubes by alloreactive cytotoxic T cells and natural killer cells. J Clin Invest 1990; 86: 370.
10. Auchincloss H, Sachs DH. Xenogeneic transplantation. Ann Rev Immunol 1998; 16: 433.
11. Guerette B, Asselin I, Skuk D, et al. Control of inflammatory damage by anti-LFA-1: increased success of myoblast transplantation. Cell Transplant 1997; 6 ( 2): 101.
12. Maier S, Tertilt C, Chambron N, et al. Inhibition of natural killer cells results in acceptance of cardiac allografts in CD28-/- mice. Nat Med 2001; 7 ( 5): 557.
13. Lanier L. NK cell receptors. Ann Rev Immunol 1998; 16: 359.
14. Cretney E, Degli-Esposti MA, Densley EH, et al. M144, a murine cytomegalovirus (MCMV)-encoded major histocompatibility complex class I homologue, confers tumor resistance to natural killer cell-mediated rejection. J Exp Med 1999; 190 ( 3): 435.
15. Ramey JW, Booker SS, Kanbour-shakir A, et al. Inability to establish ectopic endometrium in a natural killer cell-deficient murine model. Immunologic, histologic and histochemical assessment. J Reprod Med 1996; 41 ( 11): 807.
16. Spencer MJ, Walsh CM, Dorshkind KA, et al. Myonuclear apoptosis in dystrophic mdx muscle occurs by perforin-mediated cytotoxicity. J Clin Invest 1997; 99 ( 11): 1.
17. Taylor MV, Ward B, Schatzle J, et al. Perforin- and Fas-dependent mechanisms of natural killer cell-mediated rejection of incompatible bone marrow cell grafts. Eur J Immunol 2002; 32 ( 3): 793.
18. Han WR, Zhan Y, Murray-Segal LJ, et al. Prolonged allograft survival in anti-CD4 transgenic mice: lack of helper T cells compared with other CD4 deficient mice. Transplantation 2000; 70 ( 1): 168.
19. Robertson TA, Grounds MD, Papadimitriou JM. Elucidation of aspects of murine skeletal muscle regeneration using local and whole body irradiation. J Anat 1992; 181: 265.
20. Musaro A, McCullagh K, Paul A, et al. Localized Igf-1 transgene expression sustains hypertrophy and regeneration in senescent skeletal muscle. Nat Gen 2001; 27 ( 2): 195.
21. Liu J, Morris MA, Nguyen P, et al. Ly491 NK cell receptor transgene inhibition of rejection of H2b mouse bone marrow transplants 1, 2. J Immunol 2000; 164: 1793.
22. Guerette B, Skuk D, Celestin F, et al. Prevention by anti-LFA-1 of acute myoblast death following transplantation. J Immunol 1997; 159: 2522.
23. Smythe GM, Hodgetts SI, Grounds MD. Immunobiology and the future of myoblast transfer therapy. Mol Ther 2000; 1 ( 4): 303.
24. Bendelac A, Rivera M, Park S-H, et al. Mouse CD1-specific NK1 T cells: development, specificity, and function. Ann Rev Immunol 1997; 15: 535.
25. Pimorady-Esfahani A, Grounds MD, McMenamin PG. Macrophages and dendritic cells in normal and regenerating murine skeletal muscle. Muscle Nerve 1997; 20 ( 2): 158.
26. Josien R, Heslan M, Soulillou JP, et al. Rat spleen dendritic cells express natural killer cell receptor protein 1 (NKR-P1) and have cytotoxic activity to select targets via a Ca2+-dependent mechanism. J Exp Med 1997; 186 ( 3): 467.
27. McMahon CW, Raulet DH. Expression and function of NK cell receptors in CD8+ T cells. Curr Opin Immunol 2001; 13: 465.
28. Boulanger A, Asselin I, Roy R, et al. Role of non-major histocompatibility complex antigens in the rejection of transplanted myoblasts. Transplantation 1997; 63: 893.
29. Qu Z, Balkir L, van Deutekom JC, et al. Development of approaches to improve cell survival in myoblast transfer therapy. J Cell Biol 1998; 142: 1257.
30. Grounds MD, White J, Rosenthal N, et al. The role of stem cells in skeletal and cardiac muscle repair. J Histochem Cytochem 2002; 50 ( 5): 589.
© 2003 Lippincott Williams & Wilkins, Inc.