New intranasal and injectable gene therapy for healthy life extension.

As the global elderly population grows, it is socioeconomically and medically critical to provide diverse and effective means of mitigating the impact of aging on human health. Previous studies showed that the adeno-associated virus (AAV) vector induced overexpression of certain proteins, which can suppress or reverse the effects of aging in animal models. In our study, we sought to determine whether the high-capacity cytomegalovirus vector (CMV) can be an effective and safe gene delivery method for two such protective factors: telomerase reverse transcriptase (TERT) and follistatin (FST). We found that the mouse cytomegalovirus (MCMV) carrying exogenous TERT or FST (MCMVTERT or MCMVFST) extended median lifespan by 41.4% and 32.5%, respectively. We report CMV being used successfully as both an intranasal and injectable gene therapy system to extend longevity. Specifically, this treatment significantly improved glucose tolerance, physical performance, as well as preventing body mass loss and alopecia. Further, telomere shortening associated with aging was ameliorated by TERT and mitochondrial structure deterioration was halted in both treatments. Intranasal and injectable preparations performed equally well in safely and efficiently delivering gene therapy to multiple organs, with long-lasting benefits and without carcinogenicity or unwanted side effects. Translating this research to humans could have significant benefits associated with quality of life and an increased health span.

The goal of achieving healthy longevity has remained a challenging subject in biomedical science. It has been well established that aging is associated with a reduction in telomere repeat elements at the ends of chromosomes (1), which in part results from insufficient telomerase activity. Importantly, the biological functions of the telomerase complex rely on telomerase reverse transcriptase (TERT) (2). TERT plays a major role in telomerase activation, which in turn, lengthens the telomere DNA (2, 3). Because telomerase supports cell proliferation and division by reducing the erosion of chromosomal ends in mitotic cells (4), animals deficient in TERT have shorter telomeres and shorter lifespans (5, 6). Recent studies on animal models have supported the therapeutic efficacy of TERT in increasing healthy longevity and reversing the aging process (7, 8). Telomere shortening also increases the risk of heart disease (9, 10). The follistatin (FST) gene encodes a monomeric secretory protein that is expressed in nearly all mammalian tissues (11). In muscle cells, FST functions as a negative regulator of myostatin, a myogenesis inhibitory signal protein. FST overexpression is known to increase skeletal muscle mass in transgenic mice by 194 to 327% (12) by neutralizing the effects of various TGF-β ligands involved in muscle fiber breakdown, including myostatin and activin inhibition complex (13). FST knockout mice have smaller and fewer muscle fibers, show retarded growth, skeletal defects, and reduced body mass, and they die within a few hours after birth. The acceleration of these degenerative trends post FST knockout underscore an important role of FST in skeletal muscle development (14). Aged mice have exhibited loss of motor unit function with impaired neuromuscular junction transmission (15). It has been shown that follistatin expression in aged mice not only increased muscle mass but also improved the neuromuscular function (16). These findings strongly implicate the therapeutic potential of FST in the treatment of muscular dystrophy, muscle loss, and impaired neuromuscular function caused by aging or microgravity. Based on this evidence and supporting assumptions, TERT and FST are among prime candidates for gene therapy protocols directed to improve healthy lifespans.

As more longevity-supporting factors are discovered, it is natural to explore potential large capacity vectors for delivering multiple genes simultaneously. Unlike adeno-associated virus (AAV), lentiviruses or other viral vectors now commonly used for gene delivery, cytomegaloviruses (CMVs) have a large genome size and unique ability to incorporate multiple genes (17). Further, cytomegaloviruses do not integrate their DNA into the host genome during the infection cycle, thereby mitigating the risk of insertional mutagenesis (17). CMV infections are usually asymptomatic in most healthy adults, but can become problematic in neonates or transplant patients (18, 19). Human CMV (HCMV) has been proven to be a safe delivery vector for expressing therapeutic proteins in human clinical trials (20). Mouse CMV (MCMV) and HCMV are similar in many aspects, including viral pathogenesis, homology, viral protein function, viral gene expression, and viral replication (21, 22). Cytomegalovirus vector has been proven to be a potent delivery vector for delivering foreign genes and is utilized in different immunotherapies, including cancer (23), tuberculosis (TB) (24), acquired immunodeficiency syndrome (AIDS) (25, 26), malaria (27), and many others. Using MCMV as a viral vector, we examined the therapeutic potential of TERT and FST gene therapy to offset biological aging in a mouse model, and demonstrated significant lifespan increase, as well as positive metabolic and physical performance effects. We believe further studies may elucidate the full CMV cargo capacity and effectiveness. Translational studies are required to determine whether our findings can be replicated in human subjects.

Results

Construction of MCMVTERT and MCMVFST.

We developed an MCMV vector that expresses luciferase as a reporter gene (MCMVLuc) to easily monitor MCMV infection and cellular replication in cell culture and in a mouse model. MCMVLuc replicated as well as its parental virus, and we used it throughout this study as an empty viral vector control or wild-type MCMV virus inoculation (WT) (28).

We constructed recombinant MCMV vectors expressing FLAG-tagged genes TERT and FST genes (MCMVTERT and MCMVFST) and demonstrated that they replicated as productively as MCMVLuc (WT) in mouse fibroblast cells and in vivo (Fig. 1 ). Mouse TERT was fused with C terminus FLAG-tag via two intervening amino acids (threonine and arginine).

Fig. 1.

Construction and verification of MCMVTERT and MCMVFST. (A) TERT-3′ and FST-3′ FLAG constructs. (B) Expression of TERT (∼131 kDa) or FST (∼41 kDa) proteins in MCMVTERT- or MCMVFST-treated NIH/3T3 cells. (C) PFU assay growth curve of MCMVLUC (WT), MCMVTERT, and MCMVFST in NIH/3T3. n = 3 per group. Total photon counts are represented in log10 scale. Data are presented as mean ± SEM. (D) Luciferase signal in vivo 3 d after IP inoculations with mock, WT, MCMVTERT, and MCMVFST. (E) Detection of TERT by ELISA in serum of treated 8-mo-old mice over 1 mo. Two-way ANOVA with Tukey’s posttests. P < 0.001 TERT-IN vs. WT-IN group at the same time point; P < 0.001 TERT-IP vs. WT-IP group at the same time point. n = 3 per group. Data are presented as mean ± SEM.

To determine the expression kinetics of target protein and therapeutic efficacy of CMV as a delivery vector, we evaluated TERT protein levels in the blood of 8-mo-old mice over a period of 1 mo posttreatment. TERT protein expression delivered intraperitoneally (IP) or intranasally (IN) peaked at 7 d and then gradually decreased, reaching the basal level at around day 25 (Fig. 1), confirming the vector’s ability to deliver exogenous proteins in vivo. This was further confirmed by the analysis of TERT and FST mRNAs and proteins in blood and tissues harvested from treated animals, as described below.

Significant Lifespan Extension.

Seven groups of nine aged female C57BL/6J mice received mock (IP), WT-IN, WT-IP, MCMVTERT-IN, MCMVTERT-IP, MCMVFST-IN, and MCMVFST -IP, respectively, at doses of 1 × 105 plaque-forming units (PFU). Treatment started in 18-mo-old mice, equivalent to ∼56 y olds in humans (Fig. 2) (29). One mouse per group was killed at 24 mo for tissue analyses, while the remaining subjects were monitored for physical and physiological changes until their natural death. The administration of respective viruses in each group was discontinued when all mice died in the control group (29 mo). The administration of recombinant viruses was resumed at 32 mo to check the effect of the antiaging therapy on maximum lifespan.

Fig. 2.

MCMVTERT and MCMVFST significantly extend lifespan. (A) Survivorship curve comparison, eight mice per group. The survival curve of mice in each group was determined by a Kaplan–Meier survival curve. χ2 test, P < 0.001 TERT-IP vs. WT-IP and TERT-IN vs. WT-IN group at the 50% survival probability; P < 0.001 FST-IP vs. WT-IP and FST -IN vs. WT-IN group at the 50% survival probability. n = 8 per group. (B) C57BL/6J mice and human age equivalence at the start of experimental treatment. (C) TERT and (D) FST proteins by ELISA in blood serum from 24-mo-old mice. Two-way ANOVA with Tukey’s posttests. P < 0.001 TERT-IN vs. WT-IN group at the same time point; P < 0.001 TERT-IP vs. WT-IP group at the same time point. n = 3 per group. Data are presented as mean ± SEM. (E) The fold increase of TERT and FST mRNA levels in organs of MCMVTERT- and MCMVFST-treated mice by RT-qPCR in comparison to WT-treated mice. n = 3 per group. Data are presented as mean ± SEM.

Mock and WT controls died at 26.7, 26.5, and 26.4 mo (median), consistent with previous reports on lifespan of female C57BL/6J mice (30, 31). The median age at death in MCMVFST-treated groups was 35.1 mo (32.5% increase), while MCMVTERT-treated mice lived 37.5 mo (41.4% increase) (Fig. 2 and ). A maximum lifespan of 41.2 mo was observed for MCMVTERT-treated mice, which is superior to previous reports (7, 32). In the case of MCMVFST, a maximum lifespan of 38.0 mo was observed. This result exceeds the longevity achieved with a single dose of AVV9-TERT in the same animal model (13 to 24% when delivered in a single dose in 2- and 1-y-old mice, respectively) (7). Interestingly, CMV therapy was equally effective regardless of route of inoculation, although the mechanism of dissemination differs, suggesting that expression of the therapeutic load is not substantially affected by the vector’s interaction with the immune system (33).

Systemic TERT and FST Expression.

The amounts of TERT (Fig. 2) or FST (Fig. 2) proteins measured by enzyme-linked immunosorbent assay (ELISA) in serum increased daily in the first 4 d postinoculation, while endogenous protein levels remained largely unchanged in the control groups. The levels of mRNAs of TERT and FST determined by RT-qPCR in brain, heart, kidney, liver, lung, and skeletal muscle from MCMVTERT or MCMVFST mice were 1.9 to 7.8 times greater than WT-treated controls in all tested organs (Fig. 2). The variations of the mRNA levels of TERT or FST in different tissues may be due to the different tropism of CMV and the posttranscriptional modification of TERT and FST.

Telomere Length Increased in TERT-Treated Mice.

We determined the telomere length in kidney and muscle tissues in one mouse from each group using the quantitative fluorescence in situ hybridization (Q-FISH) method as described previously (34, 35). Q-FISH results demonstrated an increase in telomere length in kidney of 3.1 times, as compared to the untreated group (Fig. 3 ). We observed that the number of longer telomeres was higher in the kidney of TERT-treated mice as compared with untreated or WT-treated mice (Fig. 3 ). Telomere length was not changed significantly in FST- or WT-treated mice, suggesting a TERT-specific effect in the TERT group (Fig. 3 ). We also determined the telomere length in skeletal muscle of TERT-treated mice. We observed that the telomere length in muscle of TERT-treated mice was increased approximately threefold as compared with untreated mice (Fig. 3 ). Telomere length was not increased in muscles of WT- or FST-treated mice. We also determined the telomere length in different organs using a real-time qPCR. We did not see a variation in telomere length in different tissues of TERT-treated mice. We believe this could be due to a minimum TERT expression level required to maintain telomere length. MCMV was able to deliver TERT to different organs. The relative telomere length in heart, liver, kidney, brain, lung, and muscle in 24-mo-old MCMVTERT-treated mice was approximately three times greater than in control mice of the same age, and only ∼8% shorter than an 8-mo-old control (Fig. 3). An increase in telomere length in various organs of the MCMVTERT-treated group indicates that functional TERT was delivered by the CMV vector.

Fig. 3.

MCMVTERT-treated mice have longer telomeres. Telomere-FISH images of kidney (A) and muscle (B) tissue sections from 24-mo-old mice in indicated groups. Sections were stained with a CY3-labeled peptide nucleic acid probe complementary to telomeric repeats. The images show that TERT-treated mice have higher telomere fluorescence signal intensities compared to mice in the other groups. Telomeres, red; DAPI, blue. The mean fluorescence signal intensities in quantification of telomeres in the image A of kidney (C) and image B of muscle (D) tissue sections of treated mice in indicated groups. The error bars show SD. (E) Relative telomere lengths in organs from 24-mo-old mice vs. an 8-mo-old control was determined by RT-qPCR. The 36B4 gene was used for normalization (57). The relative telomere length was calculated by ΔCT value as described previously (58). Two-tailed unpaired t test. ***P < 0.001, TERT-IN vs. WT-IN group; P < 0.05, P < 0.01 FST-IN vs. WT-IN group. n = 3 per group. Data are presented as mean ± SE. NS, not significant.

Hair and Weight Loss Prevention.

MCMVTERT- and MCMVFST-treated animals retained coat shine and smooth texture, as well as experiencing less hair loss when compared to mock- or WT-infected mice (Fig. 4). This finding correlates with previous reports that TERT expression in skin facilitates hair growth by enhancing the follicle stem cell proliferation (36), and that follistatin has an important function in maintaining healthy skin and hair in old mice (37, 38). Body weight peaked at 23 mo for all treatment groups except for MCMVFST mice, whose weights continued to increase until 27 mo to ∼33% heavier than the age-matched mock and WT controls. MCMVTERT treatment also showed less weight loss over time compared to the mock and WT groups (Fig. 4). Administration of MCMVTERT and MCMVFST was interrupted after mice reached 29 mo of age when all mice in the control groups died (Fig. 4, red arrow) but was resumed at 32 mo (Fig. 4, green arrow). When the treatment was stopped, the weights of MCMVTERT and MCMVFST groups declined, but the rate of weight loss decreased immediately upon therapy reinitiation. Future studies would be of interest to determine whether an uninterrupted monthly administration has a different outcome in longevity extension.

Fig. 4.

MCMVTERT and MCMVFST dramatically improved physical and physiological conditions. (A) Hair and body appearance after 8 mo of treatment. (B) Biweekly body weight averages of surviving mice in each group. Treatment interruption (red arrow) and reinitiating (green arrow). n = 8 per group. Data are presented as mean ± SEM. (C) Average number of climbing attempts in 3 min. Two-tailed unpaired t test. P < 0.001 TERT-IP vs. WT-IP and TERT-IN vs. WT-IN group. n = 3 per group. Data are presented as mean ± SEM. (D) Beam crossing average execution time. Two-tailed unpaired t test. P < 0.001 TERT-IP vs. WT-IP and TERT-IN vs. WT-IN group; P < 0.001 FST-IP vs. WT-IP and FST -IN vs. WT-IN group. n = 3 per group. Data are presented as mean ± SEM. (E) Glucose tolerance test. Two-way ANOVA with Tukey’s posttests. P < 0.001 TERT-IP vs. WT-IP and TERT-IN vs. WT-IN group at the same time point; P < 0.001 FST-IP vs. WT-IP and FST -IN vs. WT-IN group at the same time point. n = 3 per group. Data are presented as mean ± SEM. (F) HbA1c levels in mock-, WT-, MCMVTERT-, and MCMVFST-treated mice. Two-tailed unpaired t test. P < 0.001 TERT-IP vs. WT-IP and TERT-IN vs. WT-IN group; P < 0.001 FST-IP vs. WT-IP and FST -IN vs. WT-IN group. n = 3 per group. Data are presented as mean ± SEM.

Improved Activity and Motor Coordination.

MCMVTERT-treated animals were ∼40% more active than control mice in attempts to escape in a beaker test that was performed by 24-mo-old mice (Fig. 4). Additionally, mice (24 mo old) treated with MCMVTERT or MCMVFST completed a beam-crossing coordination test in ∼7.5 s and 12.5 s, respectively, as opposed to the controls (∼43 s), demonstrating superior coordination (Fig. 4).

Increased Glucose Tolerance.

Glucose tolerance is known to decrease with aging. Here, we used a glucose tolerance test in fasted mice (22 mo old) from each treatment group (Fig. 4) (39). The average peak glucose concentration was ∼33% lower for TERT and ∼28% lower for FST treatments than for controls. Moreover, blood sugar levels reached baseline at 3 h postadministration in MCMVTERT- and MCMVFST-treated mice, in contrast to ∼8 h for control mice. In addition, the level of glycated hemoglobin (A1C) in treated mice (23 mo old) was 4.5% (TERT) and 4.7% (FST), versus mock (7.9%) or WT (8.8%) (Fig. 4). TERT and FST treatments were equally effective in blood glucose processing.

Mitochondrial Integrity in Muscle.

Mitochondria provide the essential metabolic support for an organism, and therefore play a central role in both lifespan determination and cardiovascular aging (40, 41). Here, we killed 24-mo-old mice from mock-, WT-, MCMVTERT-, and MCMVFST-treated groups and sectioned heart and skeletal muscle tissues to examine subcellular structures of cardiomyocytes and skeletal muscle cells by electron microscopy. The number of mitochondria with connected cristae and mitochondrial area in the aged cardiomyocyte (Fig. 5), as well as within cells of skeletal muscle (Fig. 5) of MCMVTERT- or MCMVFST-treated mice were comparable to 6-mo-old mouse controls and substantially better than in age-matched control mouse tissues. These results suggest that MCMVTERT and MCMVFST preserved mitochondrial structure and sustained mitochondrial biogenesis.

Fig. 5.

MCMVTERT and MCMVFST prevent mitochondrial deterioration in mice. (A and B) Representative EM images from the heart and skeletal muscle of untreated young mice and 24-mo-old mice treated with mock, WT, MCMVTERT, and MCMVFST (IN groups are shown). (Scale bar, 500 nm.) Quantitative analyses of the number of mitochondria with connected cristae and mitochondria area are on the Right of A and B. Two-tailed unpaired t test. ***P < 0.001 TERT vs. WT group; ###P < 0.001 FST-IN vs. WT group. NS, not significant. n = 20 per group. Data are presented as mean ± SEM.

Discussion

Aging is often accompanied by the development of chronic conditions. The socioeconomic burden imposed by the diseases of aging could be lessened by maintaining a healthy aged population. In this exercise, we explored an approach to achieve healthy aging and illustrate that CMV can be used as a monthly inhaled gene therapy for delivering the exogenous genes TERT or FST safely and effectively. Interestingly, CMV therapy was equally effective regardless of route of inoculation, even though the mechanism of dissemination differs, which suggests that expression of the therapeutic load might not be substantially affected by the vector’s interaction with the immune system. Our results are congruent with other studies, which demonstrated that the olfactory route is a preferred natural route of CMV entry in murine models (33). Because herpesviruses are ubiquitous and acquired early, it has been proposed that a mutualistic relationship has developed between the coevolving host and virus, where the latter even offers some immunomodulatory and homeostatic advantages to the hosts when in equilibrium (42). CMV, like some AAV serotypes, has broad tissue tropism, but has an advantage over AAV as a delivery vector due to its ample cargo capacity (43–45). CMV’s large genome accommodates nearly 75% dispensable genes, many of which are involved in immune evasion mechanisms that protect it from aggressive viral clearance responses regardless of route of entry (45, 46). Olfactory infection spreads through dendritic cells, which migrate to lymph nodes and then extravasate into the bloodstream, whereas IP inoculation is expected to engage a wider range of myeloid progenitor cells, which in turn dictates viral dissemination and immune response outcomes (47).

FST-treated mice showed an increase in body mass. Other studies reported body weight increase upon follistatin administration in animal models and demonstrated that it specifically correlated with increased muscle mass (16, 48–51). The increased robustness may explain the improved motor control in executing the beam test. However, it was unexpected that CMV-based FST gene therapy alone would increase longevity to the extent observed. Although it is known that FST has a concentration-dependent inhibiting effect on the myostatin-driven rate of muscle breakdown, which contributes to increasing frailty in aging individuals, the overall effect of increasing longevity warrants further inquiry. We anticipate that sarcopenia, muscular dystrophy, or even special circumstances causing muscle atrophy, such as low-gravity exposure during space travel (52), could be mitigated with a CMV-based FST gene delivery method.

Another surprising finding was the equivalent effectiveness of both treatment regimens in blood glucose control, because the cellular mechanisms activated by TERT and FST, which ultimately result in glucose control, are different. FST has a systemic role in up-regulating factors controlling mitochondrial biogenesis, energy metabolism, cellular respiration, and thermogenesis, inducing browning of white adipose tissue (53). The FST interference with the TGF-beta signaling pathway resulted in the efficient regulation of glucose homeostasis we observed. On the other hand, TERT seems to act at the level of pancreatic beta cells by up-regulating insulin secretion rather than effecting glucose uptake (54). Nonetheless, telomerase is known to interact with various cellular inflammatory pathways to reduce oxidative stress and has been detected in mitochondria, where it protects mitochondria from oxidative damage; we believe this explains the systemic benefits and increased longevity (10). Furthermore, FST and TERT have clear positive effects in neurological diseases, and the fact that our treatment clearly showed that brain FST and TERT levels increase significantly over baseline, supports its use for treatment of these conditions (55, 56). It would be of great interest to understand the compounded effect that these two therapies might have when delivered simultaneously.

Finally, our therapeutic regimen appeared to require monthly administration to have continuous effects. This may be advantageous when treatment indications do not require permanent expression of therapeutic load, but rather episodic or during specific circumstances, to achieve a reduced risk of long-term consequences in case of adverse reactions, should any occur.

In summary, our study justifies further efforts to investigate the use of CMV TERT and FST vectors against aging-related chronic inflammatory conditions, type 2 diabetes, sarcopenia, dementia, lung, kidney, and heart diseases responsible for decreased quality of life and premature death.

Materials and Methods

Cells and media, construction of MCMV with a luciferase marker, construction and characterization of recombinant MCMV vectors, measurement of telomere length, measurement of TERT and FST expression in different tissue, determination of TERT and FST protein levels in the sera, body weight, and body hair analyses, activity test and beam coordination test, glucose tolerance and glycosylated hemoglobin A1c (HbA1c) test, transmission electron microscopy, and statistical analyses can be found in . Approved Institutional Biosafety Committee and Institutional Animal Care and Use Committee protocols were followed as per guidance by Rutgers University, New Jersey.