Targeted CRISPR regulation of ZNF865 enhances stem cell cartilage deposition, tissue maturation rates, and mechanical properties in engineered intervertebral discs

Abstract

Cell and tissue engineering based approaches have garnered significant interest for treating intervertebral disc degeneration and associated low back pain due to the substantial limitations of currently available clinical treatments. Herein we present a clustered regularly interspaced short palindromic repeats (CRISPR)-guided gene modulation strategy to improve the therapeutic potential of cell and tissue engineering therapies for treating intervertebral disc disease. Recently, we discovered a zinc finger (ZNF) protein, ZNF865 (BLST), which is associated with no in-depth publications and has not been functionally characterized. Utilizing CRISPR-guided gene modulation, we show that ZNF865 regulates cell cycle progression and protein processing. As a result, regulating this gene acts as a primary titratable regulator of cell activity. We also demonstrate that targeted ZNF865 regulation can enhance protein production and fibrocartilage-like matrix deposition in human adipose-derived stem cells (hASCs). Furthermore, we demonstrate CRISPR-engineered hASCs ability to increase GAG and collagen II matrix deposition in human-size tissue-engineered discs by 8.5-fold and 88.6-fold, respectively, while not increasing collagen X expression compared to naive hASCs dosed with growth factors. With this increased tissue deposition, we observe significant improvements in compressive mechanical properties, generating a stiffer and more robust tissue. Overall, we present novel biology on ZNF865 and display the power of CRISPR-cell engineering to enhance strategies treating musculoskeletal disease.

Statement of significance:

This work reports on a novel gene, ZNF865 (also known as BLST), that when regulated with CRISPRa, improves cartilagenous tissue deposition in human sized tissue engineering constructs. Producing tissue engineering constructs at human scale has proven difficult, and this strategy presents a broadly applicable tool to enhance a cells ability to produce tissue at these scales, as we saw an ~8–88 fold increase in tissue deposition and significantly improved biomechanics in large tissue engineered intervertebral disc compared to traditional growth factor differentiation methods. Additionally, this work begins to elucidate the biology of this novel zinc finger protein, which appears to be critical in regulating cell function and activity.

1. Introduction

The intervertebral disc (IVD) is an avascular fibrocartilaginous tissue comprised of a nucleus pulposus (NP) and annulus fibrosus (AF). The NP is a gelatinous isotropic tissue comprised predominantly of proteoglycans, type II collagen, and water [1-3]. In contrast, the AF is a highly organized anisotropic tissue comprised primarily of aligned type I collagen layered in concentric lamellae [1-3]. Together these tissues allow healthy IVDs to bear load, provide flexibility to the spine, and act as shock absorbers [2]. However, degenerative disc disease (DDD) leads to a breakdown of the extracellular matrix (ECM) components and organization [4], senescent cells [4,5], and inflammation [6-8], which results in compromised mechanical function [4] and nociceptor sensitization [9-11], which are often associated with low back pain (LBP) [1,2,4].

LBP and associated DDD are significant health concerns throughout the world. It is estimated that 80 % of individuals will experience an episode of LBP in their life, with nearly 40 % having an additional recurrence of LBP within 1-year [12]. A major contributing factor to LBP is DDD [4,13,14]. Twenty percent of teens exhibit signs of mild degeneration, and subsequently 60 % of discs in 70-year-olds exhibit signs of severe degeneration [4,14]. Mild cases of DDD are generally treated palliatively, however, in severe cases the treatment options are limited to invasive surgical interventions, including spinal fusion or total artificial disc replacement [4,15-17]. Unfortunately, these interventions severely limit the natural function of the spine and have high rates of failure [18,19].

Tissue engineering/cell therapy offers the potential to overcome limitations associated with current treatment strategies for DDD and associated LBP by replacing the diseased tissue with functional tissue. One such example is composite tissue-engineered IVDs (e.g. disc-like angle ply structures (DAPS)), which have been investigated both in vitro and in vivo and have shown promise as a therapeutic option to treat DDD [20-24]. However, as the size these constructs increase and approach human physiologic scale; tissue deposition, construct maturation, and cell viability suffer as a result of increased diffusion distances, severely limiting the outcomes of such therapies [20,24]. These concerns can be attributed to the avascular and complex microenvironment within the scaffold, similar to the native IVD, as the size increases to human physiologic relevancy [1,4,25,26], which severely limits the efficacy (i.e. matrix production and cell viability) of cells seeded within such scaffolds. However, recent advances in cell engineering may provide an ability to engineer cell phenotype and tightly regulate matrix production, which has high potential to overcome these challenges.

Over the last decade, clustered regularly interspaced short palindromic repeats (CRISPR)-guided gene modulation has been used to engineer and control cell phenotype [7-11,27-36]. Briefly, this system is comprised of three main parts, a 20 base pair guide RNA (gRNA) which targets the promoter of the gene of interest, a deactivated endonuclease dCas9 which complexes with the gRNA to target the promoter, and an effector molecule (e.g. VPR), which is fused to the dCas9 to drive upregulation of the targeted gene [27,29,31]. This system, called CRISPR-activation (CRISPRa), is highly effective in controlling cell phenotype and directing stem cell differentiation [29,31,37,38].

Previously, we have shown that targeted multiplex CRISPRa upregulation of the genes aggrecan (ACAN) and collagen-II (Col2a1) can drive synergistic differentiation in human adipose-derived stem cells (hASCs) without the use of growth factors [33]. The ability to control cell phenotype and guide differentiation without growth factors overcomes the challenges of maintaining cell phenotype once delivered into complex in vivo environments, which may not be conducive to the desired cell phenotype [39-41]. The ability to perform multiplex gene modulation provides potential to further engineer cells for these environments, and gene targets have great promise to improve the efficacy of cell therapies broadly.

Recently, our lab discovered a zinc-finger (ZNF) protein, ZNF865 (also known as BLST [42]), that has shown potential as a tool to enhance cell and tissue engineering therapies. Briefly, ZNFs represent the largest group of regulatory proteins within eukaryotic genomes and due to their extensive interactions, ZNFs regulate diverse cellular processes, such as transcription, cell migration, growth, proliferation, and differentiation [43,44]. We have previously shown that ZNF865 regulates cellular senescence, cell cycle, DNA replication, and protein processing and could be used as a gene therapy to rescue cell populations from senescence [45]. In addition, we believe that CRISPRa regulation of ZNF865 could be used in cell therapy applications to enhance cell proliferation and protein production rates. We hypothesize that targeted regulation of ZNF865 in addition to ACAN and Col2a1 in hASCs can be used in cell and tissue engineering applications to increase tissue deposition and scaffold maturation rates as DAPS are scaled up to human-sized tissues.

Herein we present ZNF865 as a tool to enhance cell and tissue engineering therapies. Initially, we show that multiplex regulation of ACAN/Col2a1/ZNF865 increases the protein production rate of hASCs while not directly altering the healthy chondrogenic phenotype: high expression of aggrecan and collagen II with no expression of collagen X. Subsequently, we investigate changes in proliferation rates and expression of ACAN and Col2a1 resulting from ZNF865 regulation. In a pellet culture model, we show that the addition of ZNF865 increases overall tissue deposition and retention of key cartilage matrix proteins compared to just ACAN/Col2a1 regulation alone. From these results, we investigated the efficacy of our CRISPR-edited hASCs in medium and large human-sized DAPS and show ACAN/Col2a1/ZNF865 hASCs increase overall matrix deposition, tissue maturation rate, and mechanical properties of DAPS. Taken together, our results present ZNF865 as a powerful tool to enhance cell and tissue engineering therapies treating musculoskeletal disease.

2. Materials and methods

2.1. Experimental overview

ZNF865 was investigated as a tool to boost the efficacy of cell and tissue engineering therapies (Fig. 1). CRISPRa was used to multiplex upregulate ACAN, Col2a1, and ZNF865 in human adipose-derived stem cells (hASCs) and investigated for changes in gene expression. Subsequently, targeted regulation of ZNF865 increases cell proliferation rates in ACAN/Col2a1 CRISPR-engineered hASCs. In addition, ZNF865 enhances cartilaginous extracellular matrix (ECM) deposition in pellet culture and tissue-engineered constructs. Finally, ACAN/Col2a1/ZNF865 CRISPR-engineered hASCs were evaluated in human-sized tissue-engineered IVDs for their ability to enhance tissue maturation, cartilage deposition, and mechanical properties.

Fig. 1. Overview of the experimental design.

The ability to engineer stem cell phenotype utilizing targeted CRISPRa upregulation of Col2a1, ACAN, and ZNF865 presents a novel tool to manipulate stem cell differentiation and boost therapeutic outcomes. RNAseq of CRISPR-engineered cells investigates the Biological Processes affected by Col2a1, ACAN, and ZNF865 upregulation. A pellet culture chondrogenic differentiation model displays the effectiveness of ZNF865 CRISPR-engineered hASCs. CRISPR-engineered hASCs enhance cartilaginous matrix deposition in tissue-engineered intervertebral disc. CRISPR-engineered discs show improved mechanical properties compared hASCs with GFs. GFs is growth factors.

2.2. General cell culture

All cell culture was performed in standard culture conditions (21 % O₂, 5 % CO₂, 37 °C), with media changes every 2–3 days. HEK 293T: Complete growth medium for cell culture for HEK 293 (ATCC CRL-1573) cells consists of HG-DMEM (ThermoFisher Scientific, Waltham, MA) supplemented with 10 % fetal bovine serum (FBS) (ThermoFisher Scientific), 25 μg/mL gentamicin (Corning), and 25 mM HEPES (ThermoFisher). hASC: Complete growth medium for culturing of human adipose-derived stem cells (hASCs, ATCC SCRC-4000) consists of Lonza ADSC Basal Medium (Lonza, Lexington, MA PT-3273), 10 % MSC FBS (ThermoFisher Scientific), 5 mL GlutaMax (ThermoFisher Scientific), 30 μg/mL Gentamicin (Corning), and 15 ng/mL Amphotericin (ThermoFisher Scientific).

2.3. Cloning

pLV-U6-gRNA-UbC-DsRed-P2A-Bsr (Addgene plasmid #83919) was a gift from Charles Gersbach. Previously, our lab designed and developed lentiviral vectors targeting Col2a1 and ACAN for driving chondrogenesis without the use of growth factors in the pLV-U6-gRNA-UBC-dsRED-P2A-Bsr plasmid [33]. Subsequently, gRNAs targeting ZNF865 and a nontarget control (NTC, Table 1) that does not target the human genome, were synthesized, annealed, phosphorylated, and ligated into individual pLV-U6-gRNA-UbC-DsRed-P2A-Bsr lentiviral upregulation expression vectors. Successful gRNA insertion was verified through Sanger sequencing [7,8,33].

Table 1.

gRNA sequences for targeting the CRISPRa system to Col2a1 [33], ACAN [33], ZNF865, and a nontarget control (NTC).

Gene Target	gRNA Sequence(s)
Col2a1	gcggtagaaaggagcagcgg
ACAN	gaccgaggggcggccgacag
ZNF865 Guide 1	ccgcacaaggatggatgagt
ZNF865 Guide 2	gacgcccagagcgtgtcgcg
ZNF865 Guide 3	gaggcgggcattcaaagcgc
ZNF865 Guide 4	tcgcccaccggaatcggccc
ZNF865 Guide 5	atcctccacgccggcgcctc
ZNF865 Guide 6	acttccgcttccgggcgggc
ZNF865 Guide 7	gcacttccggtcgggccctc
NTC	tttttaatacaaggtaatct

2.4. Lentivirus production

The lenti-EF1a-dCas9-VPR-Puro (dCas9-VPR-Puro, Addgene, #99373) or gRNA plasmid DNA was used to produce lentivirus, as previously described [7,8,33]. The amplified gRNA plasmid DNA was co-transfected into HEK 293T cells with psPAX2 (Addgene, plasmid #12260) and pMD2.G (Addgene, plasmid #12259) lentiviral packaging plasmids to create a lentivirus, as previously described [7]. Briefly, HEK 293T cells were seeded at a density of 62,500 cells/cm². The following day, the lentiviral plasmids were added to the cells with Lipofectamine 2000 (ThermoFisher Scientific), following the manufacturer’s protocol. After 24-hours, the cell supernatant was discarded and replaced with fresh medium. Cell supernatant containing the lentiviral vectors was collected at 48- and 72-h, concentrated to 100X by centrifuging the collected virus at 20,000 G for 4-h and resuspending in an appropriate volume, and then virus was stored at −80 °C until use [8].

2.5. dCas9-VPR transduction

hASCs were plated at a density of 5000 cells/cm², allowed to attach overnight, and subsequently transduced with dCas9-VPR lentivirus, (1:20 dilution) in growth medium containing 4 μg/mL polybrene. hASCs were subjected to antibiotic selection (1 μg/mL) over the course of 3-days to select for dCas9-VPR expressing cells.

2.6. gRNA transduction

After dCas9-VPR transduction, dCas9-VPR expressing hASCs were transduced with gRNA targeting lentiviral vectors. ASCs were plated at a density of 5000 cells/cm² in a 24-well plate, allowed to attach overnight, and subsequently transduced with gRNA vectors (1:160 dilution) targeting ZNF865, Col2a1, ACAN, and/or a NTC, which contains a gRNA that does not target the human genome. Transduced hASCs were examined for dsRed fluorescence after 48-hours and showed near 100 % transduction efficiency.

2.7. qRT-PCR

Successfully transduced cells were analyzed for changes in ACAN/Col2a1/ZNF865 gene expression by qRT-PCR (n = 4). 72-hours posttransduction, RNA was isolated and harvested using the Quick-RNA Micro Kit (Zymo Research, R1051) and complementary DNA (cDNA) was synthesized from the purified RNA with high-capacity cDNA reverse transcription kit with RNAse inhibitor (Applied Biosystems). cDNA was then used for qRT-PCR with TaqMan gene expression assays (ThermoFisher) for ACAN (Hs00153936_m1), Col2a1 (Hs01028956_m1), and ZNF865 (Hs05052648_s1). Gapdh (Hs02786624_g1) and Beta-2-microglobulin (B2 M, Hs00187842_m1) were used as internal standards for evaluating changes in gene expression [46]. Fold-change in mRNA expression relative to VPR-NTC hASCs was calculated using the ΔΔCT method. The top-performing ZNF865 gRNA (ZNF865 guide 1) was used for the duration of the study.

2.8. RNA-sequencing

RNA-sequencing (RNAseq) was utilized to evaluate differential gene expression due to ZNF865 upregulation. ACAN/Col2a1 hASCs were analyzed for global changes in gene expression after modification with ZNF865 or NTC CRISPRa vectors. Briefly, Total RNA was isolated from samples using a Quick-RNA Kit (Zymo Research, Irvine, CA). Isolated Total RNA (100–1000 ng) was prepared using an Illumina TruSeq Stranded mRNA Library Prep Kit with PolyA Selection and samples were submitted to the High-Throughput Genomics Shared Resource Core at the Huntsman Cancer Institute for sequencing on a NovaSeq 6000, utilizing a NovaSeq Reagent Kit v1.5_150 × 150 bp sequencing with 33 million reads per sample.

Using previously described methods, data was normalized and compared to non-target control cells [7,33]. Sequencing reads were aligned to hg38 build of the human genome and reads mapping UCSC known genes were counted using featureCounts from the SubRead package [47,48]. Reads were normalized and differential analysis was performed using DESeq2 [47,49]. Enriched GO biological processes were determined from significantly regulated genes using Enrichr [50-52].

2.9. Cell proliferation quantification and cell cycle analysis

Following successful transduction, ACAN/Col2a1 hASCs (n = 4) were evaluated for cell proliferation over the course of 3–10 days. Individual cells were counted using ImageJ or a hemacytometer [53]. Briefly, dsRed fluorescing cell counts were obtained by uploading images into a stack and thresholding the images to ensure only individual cells are shown within the image. After thresholding, cell counts were obtained by analyzing particles, generating a mask, and then counting the masks generated while excluding cells on the edges of the image. In instances where cell density was too great for thresholding and visually isolating individual cells, cells were manually counted in ImageJ using the cell counter plugin [53].

For cell cycle analysis, ZNF865 and NTC edited cells were grown to confluency in a T-75 flask, lifted from the flask, and resuspended in growth medium at a concentration of 1 million cells/mL. Following the manufacturer’s instructions, 2 drops of Vybrant^™ DyeCycle^™ Violet Ready Flow^™ Reagent (ThermoFisher Scientific, R37172) was added to the suspension before being incubated at 37 °C for 30 min. Following incubation, cells were analyzed on a Cytoflex S Flow Cytometer (Beckman Coulter Life Sciences, Indianapolis, IN). Flow cytometry data was analyzed using FlowJo Flow Cytometry Software (BD Biosciences). Gates for selecting individual cells, dsRed expressing cells, and DyeCycle^™ violet expressing cells were used to isolate and analyze our cells of interest.

2.10. Pellet culture of hASCs

To evaluate extracellular matrix (ECM) deposition in our ACAN/Col2a1 edited cells, 3D pellet cultures were performed as previously described [8,33]. Briefly, pellets were formed by resuspending ACAN/Col2a1-NTC and ACAN/Col2a1-ZNF865 edited hASCs at a concentration of 1.25 million cells/mL in a serum-free growth medium, and 200 μL aliquots of the cell suspension was pipetted into individual wells of a 96-well u-bottom plate and spun at 270G for 5 min at 4 °C. Cells were allowed to contract overnight and form pellets. The following day pellets were gently lifted from the bottom of the plate. Media was changed on the pellets every 2–3 days for 21-days and supernatant was collected during every media change. After 21-days pellets were harvested and either papain digested for biochemical analysis or fixed in formalin and submitted for histological analysis, as previously described [8,33].

2.11. Macroscopic pellet imaging and size analysis

After 21-days of pellet culture, pellets were imaged (Canon Rebel T3) in their respective wells for qualitative comparison of gross pellet morphology, as previously described [8,33]. Volume was estimated by assuming the pellets were spheres, measuring the average diameter across the pellet, and calculating volume (n = 15).

2.12. Pellet culture histological analysis

To prepare pellets for staining, pellets (n = 5) were fixed in a 10 % neutral-buffered formalin solution for 24-hours, embedded in paraffin, and 5 μm sections were mounted on glass slides [8,33]. Sections for each sample were stained with alcian blue (pH 2.5; Newcomer Supply) and counterstained in Nuclear-fast red (Newcomer Supply). Briefly, slides were deparaffinized and rehydrated to distilled water, suspended in 3 % acetic acid for 3 min, suspended in alcian blue solution at room temperature for 30-minutes, washed in distilled water for 2 min, suspended in Nuclear-fast red solution for 5 min, washed in tap water, dehydrated, cleared, and coverslipped.

2.13. Pellet biochemical analysis

The total amount of collagen content contained within papain digested pellets and supernatant (n = 10) was analyzed using a modified hydroxyproline assay [8,33]. The hydroxyproline assay was performed as previously described, with the same adjustments for hydrolysis time and adjusting the pH of the oxidation buffer to 6.5 with glacial acetic acid [8,33]. The total amount of DNA and GAG content within papain digested pellets and supernatant (n = 10) was analyzed using a previously described Hoescht dye assay and a dimethylmethylene blue (DMMB) assay, respectively [8,33,54,55].

2.14. DAPS cell seeding and culture

Naïve hASCs, ACAN/Col2a1-NTC hASCs, and ACAN/Col2a1-ZNF865 edited hASCs were evaluated for cartilage deposition in vitro in medium- and large-sized DAPS over 5- or 10-weeks of culture, respectively. DAPS 10 mm in diameter and 3 mm in height for medium-sized DAPS or 16 mm in diameter and 6 mm in height for large-sized DAPS were seeded and cultured as previously described [20]. Briefly, the NP region of DAPS was formed by suspending hASCs in chemically defined media at a density of 40 × 10⁶ cells/mL mixing with molten 4 % w/v agarose (49 °C, Type VII, Sigma-Aldrich), and cast into 6-well plates to generate agarose slabs at a final density of 20 × 10⁶ cells/mL and 2 % w/v agarose gel. Sterilized biopsy punches generated NP regions of DAPS that were 3 mm in height and 5 mm in diameter (medium-sized) or 6 mm in height and 10 mm in diameter (large-sized). NP regions were cultured in isolation for 2.5 or 3 weeks for medium or large-sized DAPS, respectively, on an orbital shaker prior to combining with AF regions of DAPS. All DAPS culture was performed in chemically defined media consisting of: high glucose DMEM supplemented with 1 % PSF, 40 ng/mL dexamethasone (Sigma-Aldrich), 50 μg/mL ascorbate 2-phosphate (Sigma-Aldrich), 40 μg/mL L-proline (Sigma-Aldrich), 100 μg/mL sodium pyruvate (Corning Life Sciences, Corning, NY), 0.1 % insulin, transferrin, and selenious acid (ITS Premix Universal Culture Supplement; Corning), 1.25 mg/mL bovine serum albumin (Sigma-Aldrich), 5.35 μg/mL linoleic acid (Sigma-Aldrich), and 10 ng/mL TGF-β3 (R&D Systems, Minneapolis, MN) with media changes every 2–3 days, as previously described [20,23,24].

The AF region of the DAPS was fabricated using a 14.3 % w/v solution of poly(ε-caprolactone) (PCL) dissolved in a 1:1 mixture of tetrahydrofuran and N,N-dimethylformamide and electrospun onto a grounded rotating mandrel, as previously described [56]. Native lamellar architecture was replicated by fabricating PCL sheets that were 250 μm (medium DAPS) or 350 μm (large DAPS) thick and cut at a 30° angle to orient native fiber direction. Sheets were cut into strips 3 mm wide and 150 mm long (medium DAPS) or 6 mm wide and 153 mm long (large DAPS) which were hydrated through a gradient of ethanol and coated overnight in 20 μg/mL fibronectin (Sigma-Aldrich) in PBS. hASCs were suspended in growth medium, seeded onto PCL strips at 1.5 million or 3 million cells per side (density of 3333 cells/mm²), for medium and large DAPS, respectively, and cultured for 1 or 1.5 weeks for medium- and large-sized DAPS, respectively. The cell-seeded AF region was assembled in layers using either 2 strips (medium DAPS) or 4 strips (large DAPS) with fibers oriented in alternating directions as strips were layered (± 30°). Layered strips were then wrapped to form concentric lamellae in a custom mold, as previously described [57]. AF regions were cultured in the same chemically defined media as NP regions with media changes every 2–3 days on an orbital shaker. Following 1.5-weeks of culture around the mold, AF regions were removed from the mold and combined with NP regions, at which point combined NP and AF regions of DAPS were cultured for 2.5-weeks for medium DAPS or 7-weeks for large DAPS, resulting in a total of 5- or 10-weeks of total culture, respectively.

Medium DAPS were evaluated statistically by a Pearson’s Chi-Squared Analysis with DAPS either being considered Success or Fail. Successful DAPS were DAPS that produced sufficient ECM to maintain shape without unrolling of the AF after removal from the molds and upon combining with the NP hydrogel. Failed DAPS were DAPS that did not maintain shape, AF began to unroll, and had to be pinned together during the 5-week culture period. Subsequently, following 5 or 10-weeks of culture DAPS were subjected to mechanical testing (n = 3), biochemical analysis (n = 3), and histology (n = 2–3), as described below.

2.16. Mechanical testing

For evaluating compressive mechanical properties following 10-weeks of culture, large-sized DAPS were subjected to unconfined compression testing, as previously described [20]. Briefly, a 0.05 N pre-load was applied followed by 20 cycles of compressive loading at 0.5 Hz from 0.05 N to 48 N (0.24 MPa) and then by a creep test consisting of [20]. The stress-strain response in the 20th cycle of compression was analyzed using a bi-linear fit in MATLAB to quantify the toe and linear region moduli, transition stain, and compressive range of motion (ROM) [23]. Creep strain was calculated by normalizing displacement during the creep test to DAPS height.

2.17. DAPS biochemical analysis

Medium and large-sized DAPS GAG, collagen, and DNA content was evaluated by separating the AF and NP regions and individually digesting overnight in proteinase K at 60 °C, as previously described [20]. GAG content was determined using a DMMB assay, collagen content was determined utilizing a chloramine-T hydroxyproline assay, and DNA content was quantified using a Quanti-iT PicoGreen dsDNA assay kit (Life Technologies, Carlsbad, CA), as previously described [20].

2.18. DAPS histological evaluation

For histological assessment of matrix deposition after 5-weeks and 10-weeks of culture, DAPS were fixed in 10 % neutral buffered formalin, embedded in paraffin, and sectioned in the axial plane to 10 μm thickness. Sections were stained with hematoxylin and eosin (H&E), alcian blue, and picrosirius red.

2.19. Immunofluorescence

Immunofluorescence staining for chondroitin sulfate (DSHB 9BA12, 4 μg/mL; developed by W.M, Halfter), collagen I (Novus Biologicals, 1:100), collagen II (DSHB II-II6b3, 4 μg/mL; developed by T.F. Linsenmayer), and collagen X (DSB X-AC9, 4ug/mL; developed by T.F. Linsenmayer) was performed by serially rehydrating paraffin sections, digesting samples in Proteinase K (Dako S3020) for 5 min at room temperature, treating with Background Buster (Innovex NB306) for 30 min at room temperature, and then incubating the samples in primary antibodies overnight at 4 °C. The following day, fluorescent secondary antibodies (Invitrogen A11032, Invitrogen A11008) were applied for 1 h at 4 ° C. Before Proteinase K digestion, sections used for chondroitin sulfate staining were incubated in hyaluronidase (Sigma H3884; 18.675 units per samples) for 15 min at room temperature.

2.20. Statistics

RNA-seq statistical analysis was performed and described in their respective section (Enrichr, DESeq2, α=0.05). Statistical analysis of qRT-PCR, proliferation, and pellet biochemical data was performed using JMP Pro using a one-way analysis of variance (ANOVA) with a Tukey’s post-hoc analysis, and statistical analysis for DAPS biochemical data and mechanical testing was performed using JMP Pro with nonparametric Wilcoxon/Kruskal-Wallis test (α=0.05 for all tests).

3. Results

3.1. Upregulation of ZNF865 increases cell proliferation rates while maintaining chondrogenic phenotype

The dCas9-VPR CRISPRa system can target and differentially regulate Col2a1, ACAN, and ZNF865 (Fig. 2A). Our previous work demonstrated multiplex CRISPRa targeting of Col2a1/ACAN produced a chondrogenic phenotype in hASCs [33]. Furthermore, we have shown that CRISPR-interference (CRISPRi) can effectively block gene expression by tri-methylation of the histone, suppressing target gene expression (Supplementary Fig. 2) [7,8,11,32]. Initially, guide screening was used for targeting ZNF865. qRT-PCR was used to verify our ability to regulate ZNF865 expression and titrate ZNF865 expression up to 7.87-fold (Fig. 2B) and down to nearly 0.067-fold (Supplementary Fig. 2). Previously we have shown that targeted regulation of ZNF865 affects cell proliferation rates in various cell types [45]. This observation was conserved in hASCs where we observed a strong correlation (R² = 0.86, p = 0.0012) with increased proliferation rates correlating to increasing ZNF865 expression (Fig. 2C). When ZNF865 was downregulated, instead of increased proliferation there was cell arrest (Supplementary Fig. 2) and increased levels of senescence in various cell types [45]. In conjunction with increased proliferation rates there was a significant shift in the cell cycle for ZNF865-edited hASCs compared to NTC-hASCs (p < 0.05, Fig. 2D). In NTC-hASCs, 63 % of cells are in the G0/G1 phase and 34.17 % in the S/G2/M phase, compared to ZNF865-edited hASCs where 49.3 % of cells are in the G0/G1 phase and 47.9 % are in the S/G2/M phase. This observed shift in cell cycle was expected with the increased rates of proliferation observed in ZNF865-edited hASCs.

Fig. 2. Targeted Regulation of ZNF865 Increases Cell Proliferation in ACAN/Col2a1 hASCs.

(A) The dCas9-VPR CRISPRa system acts like a synthetic transcription factor increasing expression of target genes. (B) Initial guide screening verifies ZNF865 upregulation and displays our ability to titrate target gene regulation (n = 4, *=p < 0.05). (C) Targeted upregulation of ZNF865 results in a strong correlation of increased proliferation rates in hASCs, (D) which corresponds with a shift in cell cycle in ZNF865-engienered hASCs compared to NTC-hASCs. (E) qRT-PCR was used to monitor changes in ZNF865 expression over 8-weeks, showing increasing expression to week 4 and decreasing expression back to baseline levels by week 8 (n = 3–4, *=p < 0.05). (F) qRT-PCR verifies the upregulation of (D) Col2a1 and (G) ACAN in ACAN/Col2a1-NTC and ACAN/Col2a1-ZNF865-edited hASCs compared to naïve hASC baseline expression (n = 4, *=comparing to VPR-NTC, #=comparing to ACAN/Col2a1-NTC, *,#=p < 0.05). (H) The resulting targeted regulation of ZNF865 in ACAN/Col2a1 hASCs increases proliferation rates resulting in decreased doubling time (n = 4, *=p < 0.05). NTC is nontarget control.

Interestingly, qRT-PCR evaluated the expression of ZNF865 over several weeks and showed that peak expression of ZNF865 occurred around week 4 and week 5 and returned to baseline expression levels around week 8 (p < 0.05, Fig. 2E). Additionally, qRT-PCR verified the regulation of Col2a1 and ACAN showing significant increases in Col2a1 (p < 0.0001, Fig. 2F) and ACAN (p < 0.0001, Fig. 2G) expression compared to VPR-NTC hASCs. Previously, we showed upregulation of ZNF865 results in increased proliferation rates, a similar trend was observed in ACAN/Col2a1 CRISPR-engineered hASCs where there was a significant increase in proliferation rates in ACAN/Col2a1-ZNF865 hASCs compared to ACAN/Col2a1-NTC hASCs in monolayer. With the changes in cell proliferation rates there was a significant decrease in average cell doubling time for ACAN/Col2a1-ZNF865 hASCs compared to ACAN/Col2a1-NTC hASCs, with doubling times being 32.4 +/− 3.5 hours and 44.7 +/− 8.9 hours, respectively (p = 0.0329, Fig. 2H).

3.2. Multiplexed ZNF865 upregulation alters genes related to translation and RNA processing

RNAseq was used to evaluate global changes in gene expression due to the upregulation of ZNF865 in ACAN/Col2a1 CRISPRa upregulated hASCs. Initial analysis of the CRISPR-engineered hASCs showed two distinct populations of hASCs based on the principal component analysis (PCA, Supplementary Fig. 1). Upregulation of ZNF865 in ACAN/Col2a1 hASCs lead to the significant differential expression of 8093 genes (p < 0.05, Fig. 3A). Gene ontology (GO) analysis was performed using Enrichr with the top GO Biological Processes affected being translation, RNA processing, ribosome biogenesis, mitochondrial gene expression, gene expression, mRNA splicing, mitochondrial translation, and mRNA processing (Fig. 3B/C). A full list of all significant GO terms can be found in Supplemental Table 1. Despite these gene expression changes, ZNF865 upregulation did not affect the overall healthy chondrogenic phenotype of these cells (Fig. 3D), meaning that genes associated with healthy chondrocytes (i.e. ACAN and Col2a1) [58,59] maintain high levels of expression and genes associated with chondrogenic hypertrophy and mineralization (i.e. Col10a1, Runx2, and MMP13) [60-62] maintain low levels of expression before and after ZNF865 regulation.

Fig. 3. RNAseq evaluates global changes in gene expression related to ZNF865.

(A) Targeted upregulation of ZNF865 results in thousands of significantly differentially expressed genes. (B) The top GO Biological Processes affected by ZNF865 in ACAN/Col2a1 hASCs result in increased expression of genes related to Translation, RNA Processing, Ribosome Biogenesis, Mitochondrial Gene Expression, Gene Expression, mRNA Splicing, Mitochondrial Translation, and mRNA Processes. (C) Representative list of genes that are significantly differentially expressed genes that are associated with each term. (D) Representative transcript per million (TPM) values for ACAN, Col2a1, Col10a1, Runx2, and MMP13 in ACAN/Col2a1-ZNF865 and ACAN/Col2a1-NTC hASCs, displaying that the genes associated with a healthy chondrogenic phenotype (ACAN and Col2a1) is maintained after ZNF865 upregulation, while genes associated with chondrogenic hypertrophy (Col10a1, Runx2, MMP13) are not upregulated. NTC is nontarget control.

3.3. ZNF865 increases cartilage tissue deposition and retention in pellet culture

CRISPRa multiplexed upregulation of ACAN/Col2a1-ZNF865 (Fig. 4A) in hASCs was evaluated for cartilage ECM deposition in pellet culture. Overall, comparisons of gross pellet morphology show increased tissue deposition and pellet size in ACAN/Col2a1-ZNF865-edited hASCs compared to ACAN/Col2a1-NTC hASCs after 21-days of culture (Fig. 4B). The average volume of pellets was calculated from the images and showed significant increases in overall pellet volume in ACAN/Col2a1-ZNF865 pellets compared to ACAN/Col2a1-NTC pellets (p = 0.0001, Supplementary Fig. 3). Subsequently, pellets were stained for alcian blue to qualitatively evaluate GAG deposition, where ACAN/Col2a1-ZNF865 pellets show increased staining for GAG content compared to naive hASCs and ACAN/Col2a1-NTC pellets (Fig. 4C). Biochemical analysis shows significant increases in μg of GAG/pellet (p = 0.0006, Fig. 4D) and retention of those GAG molecules within ACAN/Col2a1-ZNF865 pellets (p = 0.0074, Fig. 4E) compared to both naïve hASCs and ACAN/Col2a1-NTC hASCs. Additionally, pellets were evaluated for collagen deposition where ACAN/Col2a1-ZNF865 pellets showed significant increases in μg of collagen/pellet (p < 0.0001, Fig. 4F) and the retention of collagen molecules within the pellets (p < 0.0001, Fig. 4G) compared to naive hASCs and ACAN/Col2a1-NTC hASCs. There was no significant difference in DNA content between groups, however there were significant increases in μg of GAG/μg DNA (p = 0.0037), and μg of collagen/μg of DNA (p = 0.0001) in ACAN/Col2a1-ZNF865 pellets compared to naïve hASCs and ACAN/Col2a1-NTC hASCs (Supplementary Fig. 3). Furthermore, ZNF865-engineered hASCs showed no mineralization in osteogenic medium, indicating that sustained upregulation of ZNF865 does not lead to osteogenesis or mineralization (Supplementary Fig. 4). These results are consistent with the RNAseq GO analysis where ZNF865 upregulation leads to increased protein production and tissue deposition of CRISPR-engineered hASCs, while not increasing expression of genes associated with chondrogenic hypertrophy or mineralization.

Fig. 4. Multiplex CRISPR-cell engineering enhances cartilage deposition.

(A) Expression cassettes for the CRISPRa system and upregulation of Col2A1, ACAN, and ZNF865. After 21-days, (B) gross pellet morphology showed increases in overall tissue deposition (scalebar = 1 mm) and increased staining for (C) alcian blue indicating increased GAG deposition (scalebar = 100 μm). Biochemical analysis showed significant increases in (D) μg GAG/pellet, (E) % GAG retained in each pellet, (F) μg collagen/pellet, and (G) % collagen retained in each pellet for ACAN/Col2a1-ZNF865 edited hASCs compared to ACAN/Col2a1-NTC and naïve hASCs without GFs (n = 5–10, *=p < 0.05). NTC is nontarget control, GFs is growth factors.

3.4. ZNF865 enhances tissue deposition in medium-sized daps

Medium-sized DAPS were seeded with ACAN/Col2a1-ZNF865 CRISPRa-edited hASCs to investigate matrix deposition in whole organ tissue engineering applications. Naïve hASCs dosed with GFs (hASCs + GFs), multiplex CRISPRa ACAN/Col2a1, or multiplex ACAN/Col2a1-ZNF865 hASCs were seeded onto AF and NP regions of DAPS and cultured for a total of 5-weeks (Fig. 5A). After 5-weeks of culture, alcian blue and picrosirius red combinatorial stain displays increases in cartilaginous tissue deposition for ACAN/Col2a1-ZNF865 hASCs compared to ACAN/Col2a1-NTC hASCs (Fig. 5B). Consistent with our previous data (Fig. 2), ZNF865-engineered hASCs displayed increased proliferation rates compared to both control groups (Fig. 5C). Out of 7 total DAPS seeded with hASCs + GFs only 1 was considered successful (i.e. the cells were able to produce sufficient matrix to adhere the AF scaffold layers into a single unit) after 5-weeks of culture. In comparison, half (3 out of 6) of the DAPS seeded with ACAN/Col2a1-NTC hASCs grew successfully, and all 9 of the DAPS seeded with ACAN/Col2a1-ZNF865 cells grew successfully (p = 0.0004, Fig. 5C). Alcian blue staining showed the same trend, with increased staining for ACAN/Col2a1-NTC and increased staining for ACAN/Col2a1-ZNF865 DAPS compared to hASCs + GFs (Fig. 5D). Biochemical analysis of DAPS showed significant increases in μg of GAG/μg of DNA in AF (p = 0.0037, Fig. 5E) and NP (p = 0.0384, Fig. 5F) regions for ACAN/Col2a1-ZNF865 DAPS compared to hASCs + GFs. Representative picrosirius red staining displays increased collagen deposition in ACAN/Col2a1-NTC and ACAN/Col2a1-ZNF865 DAPS compared to hASCs + GFs (Fig. 5G). Biochemical analysis was used to verify the significant increase in μg of collagen/μg of DNA in both the NP (p = 0.0135, Fig. 5H) and AF (p = 0.0116, Fig. 5I) of medium-sized DAPS for ACAN/Col2a1-ZNF865 DAPS compared to both the ACAN/Col2a1-NTC and hASCs + GFs groups.

Fig. 5. ZNF865 CRISPR-engineered medium-sized DAPS show increased tissue maturation rates and cartilage tissue deposition.

(A) Representative schematic showing the seeding and culture methods for medium-sized DAPS. (B) Following 5-weeks of culture ZNF865-edited DAPS show increased staining for GAG and collagen with alcian blue/picrosirius red dual stain (scalebar = 500 μm). CRISPR-engineered medium DAPS resulted in significantly more successful DAPS manufactured compared to hASCs + GFs. (C) hASCs + GFs resulted in 7 DAPS seeded, 1 deemed successful, and 6 deemed failures compared to targeted upregulation of ACAN/Col2a1-NTC resulting in 6 seeded DAPS where 3 DAPS were deemed successful and 3 deemed failures, compared to ACAN/Col2a1-ZNF865 DAPS where 9 DAPS were seeded and all 9 were deemed successful (*=p < 0.05 compared to naïve hASCs, #=p < 0.05 compared to ACAN/Col2a1-NTC). (D) Histological evaluation shows increased alcian blue staining in the NP and AF of CRISPR-engineered DAPS compared to hASCs + GFs (scalebar = 200 μm). Biochemical analysis verifies a significant increase in GAG deposition in the (E) NP and (F) AF for ZNF865-edited DAPS. (n = 4–5, *=p < 0.05). (G) Representative picrosirius red staining shows increased collagen deposition in CRISPR-engineered DAPS compared to hASCs + GFs (scalebar = 200 μm). Biochemical analysis displays a significant increase in collagen deposition in both the (H) NP and (I) AF for ZNF865-edited DAPS compared to both hASCs + GFs and ACAN/Col2a1-NTC DAPS (n = 3–5, *=p < 0.05). NTC is nontarget control GFs is growth factors.

3.5. CRISPR-engineered hASCs improve cartilage tissue deposition in human-sized daps

To evaluate the effect of CRISPR-engineered hASCs in a clinically relevant sized (human cervical disc) total disc replacement, large-sized DAPS (16 mm in diameter) were seeded and cultured for 10 total weeks with either ACAN/Col2a1-ZNF865 or hASCs + GFs and evaluated for tissue deposition histologically. Representative alcian blue and picrosirius red combination stain shows increased staining for GAG and collagen tissue deposition in CRIPSR-engineered DAPS compared to hASCs + GFs (Fig. 6A). Biochemical analysis of the NP region of DAPS shows significant increases in μg GAG/Tissue wet weight (WW) (8.5-fold, p = 0.0216, Fig. 6B), μg GAG/μg DNA (9.2-fold, p = 0.0216, Fig. 6C), μg collagen/Tissue WW (78.2-fold, p = 0.0216, Fig. 6D), and μg collagen/μg DNA (88.6-fold, p = 0.0216, Fig. 6E). However, our results show no significant difference in DNA content between groups (p = 0.66, Fig. 6F). Within the AF, there is a significant increase in μg GAG/Tissue WW (4.3-fold, p = 0.0216, Fig. 6G), μg GAG/μg DNA (1.7-fold, p = 0.0216, Fig. 6H), μg collagen/Tissue WW (5.8-fold, p = 0.0216, Fig. 6I), μg collagen/μg DNA (2.3-fold, p = 0.0216, Fig. 6J), and in DNA content (2.4-fold, p < 0.0001, Fig. 6K). Histological assessment of human-sized large-DAPS agrees with our biochemical data showing increased staining for GAG (Fig. 6L) and collagen (Fig. 6M) content in ACAN/Col2a1-ZNF865-edited DAPS compared to hASCs + GFs. Representative H&E staining of the NP revealed a healthy cell morphology located within lacunae for both groups (Fig. 6N). Noticeably, when evaluating AF region of DAPS there was increased cell infiltration into the PCL fibers in CRISPR-engineered DAPS compared to hASCs + GFs DAPS (Fig. 6O). Overall, the results presented display the efficacy of CRISPR-engineered hASCs to enhance cell infiltration and fibrocartilaginous tissue deposition for use in human-scale DAPS musculoskeletal tissue engineering applications.

Fig. 6. ZNF865-engineered Human-sized DAPS Show Dramatically Increased Cartilage Tissue Deposition.

(A) Representative alcian blue and picrosirius red dual staining shows increased staining for GAG and collagen deposition in ZNF865-edited DAPS compared to hASCs + GFs DAPS (scalebar = 3 mm). Biochemical analysis of the NP shows a significant increase in (B) μg GAG/Tissue WW, (C) μg GAG/μg DNA, (D) μg collagen/Tissue WW, and (E) μg collagen/μg DNA, however, there is no significant difference in (F) μg of DNA, for ZNF865-edited DAPS compared to hASCs + GFs (n = 3, *=p < 0.05). Within the AF, there is a significant increase in (G) μg GAG/Tissue WW, (H) μg GAG/μg DNA, (I), μg collagen/Tissue WW, (J) μg collagen/μg DNA, and (K) μg of DNA, for ZNF865-edited DAPS compared to hASCs + GFs (n = 3, *=p < 0.05). Histological assessment of human-sized DAPS displays increased staining for (L) GAG and (M) collagen in ACAN/Col2a1-ZNF865 DAPS compared to hASCs + GFs. Within the NP, (N) representative H&E images display gross cell morphology and differences in distribution between hASCs + GFs and ZNF865-engineered DAPS (scalebar = 300 μm). Within the AF, (O) representative H&E images display gross cell morphology and hASC infiltration into the PCL AF fibers in ZNF865-edited DAPS, indicating improved cell/tissue integration (scalebar = 300 μm). GFs is growth factors.

3.6. CRISPR-engineered hASCs highly express collagen II

Immunofluorescence was used to evaluate expression and distribution of chondroitin sulfate (proteoglycan), collagen I, II, and X throughout large-sized DAPS (Fig. 7A). ACAN/Col2a1-ZNF865 CRISPR-engineered DAPS show increased collagen II deposition throughout the construct with minimal expression of collagen I, compared to hASCs + GFs which shows increased expression of collagen I throughout large-DAPS (Fig. 7B). Furthermore, CRISPR-engineered DAPS show increased chondroitin sulfate (CS) deposition throughout DAPS compared to hASCs + GFs, where CS deposition is poorly distributed throughout DAPS (Fig. 7C). hASCs + GFs show increased collagen X deposition compared to ACAN/Col2a1-ZNF865 DAPS which show no collagen X deposition (Fig. 7D).

Fig. 7. CRISPR-engineered DAPS Show Increased Healthy Cartilage Deposition with Minimal Expression of Collagen I and Collagen X.

(A) Representative H&E full-disc image of Large DAPS indicating regions of interest for immunofluorescence. Representative images comparing immunofluorescence for (B) collagen I (green) and collagen II (purple), (C) chondroitin sulfate (proteoglycan, yellow)), and (D) collagen X (orange) deposition within Large DAPS.

3.7. CRISPR-engineered hASCs improve mechanics of human-sized daps

Following 10 weeks of culture, human-scale large-sized DAPS were evaluated for changes in biomechanics via unconfined compression testing (0.24 MPa, Fig. 8A). Representative stress-strain curves display a significant shift in ACAN/Col2a1-ZNF865 DAPS compared to hASCs + GFs (p = 0.0495, Fig. 8B), indicating mechanically functional tissue deposition. We observe no significant difference in the Toe Modulus (p = 0.2752, Fig. 8C) or the Linear Modulus (p = 0.8, Fig. 8D). However, there were significant differences in the Transition Strain (p = 0.0495, Fig. 8E) and Compressive Strain (p = 0.0495, Fig. 8F) in ACAN/Col2a1-ZNF865 DAPS compared to hASCs + GFs. There was no significant difference in Creep Strain (p = 0.126, Fig. 8G) between groups. Taken together, such results indicate that CRISPR-engineered DAPS increase overall tissue deposition and biomechanics, generating a stiffer tissue.

Fig. 8. ZNF865-engineered large-sized DAPS display improved mechanical properties.

(A) Following 10-weeks of culture, large-sized DAPS were subjected to unconfined compression testing utilizing 0.24 MPa. (B) Representative stress–strain curves show the shift in ACAN/Col2A1-ZNF865-edited DAPS compared to hASC + GFs DAPS, indicating the increased tissue deposition results in a stiffer overall tissue. When evaluating the moduli, the (C) toe modulus shows a non-significant increase in ZNF865-edited DAPS whereas (D) the linear modulus shows no change between groups. However, there is a significant decrease in the (E) transition strain and the (F) compressive strain in ZNF865-edited DAPS compared to hASCs + GFs. There is a non-significant increase in (G) creep strain between the two engineered discs (*=p < 0.05). NTC is nontarget control, GFs is growth factors.

4. Discussion

ZNF865 is a previously unstudied gene that has recently been shown to regulate cellular senescence, with targeted upregulation of ZNF865 showing the ability to rescue human cell populations from senescence [45]. Targeted upregulation of ZNF865 could also serve as a potential tool to enhance cell and tissue engineering therapies. Utilizing RNAseq to evaluate changes in gene expression, we showed that targeted regulation of ZNF865 affects biological processes associated with gene expression and protein processing. Subsequently, experiments were performed to evaluate ZNF865’s ability to enhance gene expression and protein production at multiple scales.

As a tool in cell and tissue engineering, ZNF865 enhanced cells’ therapeutic potential. We demonstrated that ZNF865 upregulation can be used for the rapid titratable expansion of hASCs, which is consistent with the high degree of correlation between ZNF865 expression and increased proliferation rates in CRISPR-engineered hASCs. And ZNF865-expression can be correlated to healthy cell function, noting its role as a regulator of senescence and cell function [45]. Interestingly, ZNF865 is temporally expressed over several weeks, reaching peak expression around week 4/5 and returning to baseline levels by week 8 (Fig. 2). However, its effect on cell function is maintained past 4/5-weeks of culture indicating a higher degree of regulation occurring within the cell that needs to be further investigated. In addition, we observe increased expression of Col2a1 and ACAN compared to VPR-NTC hASCs, and increased proliferation rates resulting in significantly decreased doubling times in multiplex ACAN/Col2a1-ZNF865-engineered hASCs. As a result, ZNF865 is an effective upregulation target for regulating/increasing cell production for therapeutic applications.

Further, to evaluate the efficacy of multiplex upregulated hASCs in a 3D environment, naïve hASCs, ACAN/Col2a1-ZNF865, and -NTC hASCs were evaluated in pellet culture for chondrogenic differentiation without growth factors. Previously we were able to drive chondrogenesis in CRISPR-engineered hASCs without the use of growth factors [33]. This same process was used to evaluate the capacity of ZNF865-engineered hASCs to enhance cartilage cell engineering. Multiplex ACAN/Col2a1-ZNF865 CRISPR-edited hASCs increased tissue deposition and overall size of pellets. In addition, there is an increase in GAG staining, as was expected with the increased expression of ACAN from qRT-PCR. Furthermore, our biochemical analysis displays significant increases in GAG (2-fold) and collagen (2-fold) deposition per pellet compared to control groups, as well as increased retention of these molecules compared to the controls. These results are consistent with RNAseq data that ZNF865 regulates protein processing, resulting in significantly increased protein production, indicating that ZNF865 increases both proliferation rates and protein processing. Taken together, our ability to improve cartilage ECM deposition displays functional applications in cell engineering and cell therapies.

The pellet culture experiments encouraged us to investigate the ability of ZNF865 regulation to improve outcomes in a whole tissue-engineered intervertebral disc (IVD). Recently, tissue-engineered total artificial disc replacements (ADR), such as the disc-like angle ply structure (DAPS) [20,24], are being developed as alternative treatments to the current gold-standard—spinal fusion or discectomy [63]. However, challenges with matrix deposition throughout the DAPS have been demonstrated in the scale-up to clinically relevant sizes during culture with chondrogenic growth factors [20]. We first evaluated ACAN/Col2a1-ZNF865 CRISPR-engineered hASCs in medium-sized DAPS to evaluate matrix deposition. Following 5-weeks of culture, ZNF865-edited DAPS showed robust matrix deposition throughout the scaffold compared to both the naïve control and the ACAN/Col2a1-NTC hASCs. In addition, biochemical analysis showed increased GAG (2.5-fold NP, 6-fold AF) and collagen (26-fold NP, 32-fold AF) deposition in both the NP and AF regions of DAPS made with ZNF865-edited hASCs. Further, we observed significant increases in the number of successful DAPS formed in conjunction with matrix deposition throughout the scaffold. Taken together, ZNF865-engineered hASCs can be rapidly expanded and deposit dramatically more cartilage matrix compared to naïve hASCs dosed with growth factors alone. Such results display the power of both ZNF865-engineered hASCs and CRISPR-engineered hASCs for use in tissue engineering applications.

To expand further on CRISPR-engineered hASCs improvements in tissue engineering outcomes at clinically relevant size scales, we evaluated the tissue deposition and biomechanics of CRISPR-engineered hASCs in large-sized DAPS which are similar in size to the human cervical IVD [20]. Large-sized DAPS were seeded with naïve hASCs dosed with growth factors or CRISPR-engineered ACAN/Col2a1-ZNF865 CRISPR-engineered hASCs. When evaluating large-sized DAPS histologically and biochemically, CRISPR-engineered DAPS show increased cartilage tissue deposition compared to naive hASCs dosed with growth factors alone. CRISPR-engineered DAPS display increased GAG and collagen, both histologically and biochemically in both the AF and NP. Biochemistry data demonstrated that CRISPR-engineered cells produced 88.6-fold and 5.8-fold more collagen, as well as 8.5-fold and 4.3-fold more proteoglycan in the NP and AF, respectively, compared to hASCs treated only with growth factors at this scale. Interestingly, there is no significant difference in DNA content in the NP, however, we observe a healthier cell morphology in CRISPR-engineered DAPS compared to hASCs dosed with GFs. DNA content was consistent between the pellet culture and DAPS experiments within the NP, where there was no significant difference in DNA content. However, there was a significant increase in DNA content for CRISPR-engineered DAPS in the AF region. With observing increased proliferation rates in monolayer and no significant differences in DNA content between pellets, there may be some signaling differences going on between the NP and AF of large-sized DAPS that is causing the observed differences between pellet culture and DAPS. The DNA differences may be caused by differences in composition and stiffness of the seeding substrate due to differing substrates causing stem cells to respond and differentiate accordingly [64]. The agarose NP may result in more robust differentiation, whereas the PCL fibers of the AF may result in a mixture of differentiation and proliferation [64]. However, with the increased proliferation rates in CRISPR-engineered hASCs, there is more robust cell infiltration into the polymeric AF of DAPS. The increased cell infiltration into the PCL fibers of the AF, which could be the result of increased proliferation rates, indicates a higher degree of integration compared to naïve hASCs dosed with GFs. This increased infiltration could improve functional tissue maturation and integration rates which could in turn result in a more effective tissue-engineered ADR. However, further experiments will be needed to elucidate these answers.

The targeted regulation of ZNF865 displays a high degree of control over cell cycle and cell proliferation rates, as well as protein processing and protein deposition for use in cartilage tissue engineering. As size of construct increased so did the degree of tissue deposition, with 2-fold increase in pellet culture for collagen and proteoglycan deposition to an 88.6-fold increase in collagen deposition and 8.5-fold increase in proteoglycan deposition in human-sized DAPS. The addition of ZNF865-upregulation has increased the protein processing rate of the engineered hASCs resulting in increased rates of ECM tissue deposition. With the overall increase in tissue deposition the tissue being deposited is highly specific to the targets within the CRISPR-engineered hASCs. Histological assessment shows increased staining for collagen and proteoglycan, and these results are further supported with immunofluorescence data. The immunofluorescence staining shows increases for both collagen II and chondroitin sulfate (proteoglycan) throughout the NP and AF of human-sized DAPS compared to hASCs + GFs (Fig. 7). These results are consistent with RNAseq and qRT-PCR where there is high expression of aggrecan and collagen II, displaying the specificity and efficacy of CRISPR-guided gene modulation.

Furthermore, within RNAseq data there was no expression of genes associated with chondrogenic hypertrophy (collagen X, MMP13, RUNX2), and these results are consistent in the immunofluorescence data. CRISPR-engineered hASCs show no collagen X deposition compared to hASCs + GFs which show increased collagen X deposition, indicating those cells are undergoing chondrogenic hypertrophy and mineralization, a fate associated with stem cells differentiated with TGF-β [60,62,65]. To further evaluate the inability of ZNF865-engineering to undergo osteogenesis and mineralization, ZNF865-engineered hASCs were subjected to osteogenic differentiation culture conditions and displayed no predisposition to osteogenesis or mineralization (Supplementary Fig. 4). Additionally, previous studies have shown that stem cells expressing collagen X in vitro, when implanted in vivo, resulted in calcium deposition and mineralization, whereas stem cells not expressing collagen X in vitro did not produce calcium in vivo [66]. Thus, CRISPR-engineered hASCs not depositing collagen X in vitro are unlikely to undergo mineralization in vivo showing promise for the engineered hASCs as a chondrogenic differentiation method that avoids chondrogenic hypertrophy. Taken together, our CRISPR-engineered hASCs show no propensity to undergo chondrogenic hypertrophy or mineralization in vitro indicating that the healthy chondrogenic phenotype would be maintained in vivo. Ultimately, these studies show that ZNF865-CRISPR-engineering does not hinder the chondrogenic propensity of CRISPR-engineered ACAN/Col2a1 hASCs but enhances the ability of hASCs to deposit aggrecan and collagen II and retain those molecules. As a whole, these results display the power of CRISPR-cell engineering to better control and enhance the chondrogenic differentiation outcomes of stem cells without the complication of hypertrophy and mineralization.

To evaluate the biomechanics of the human-sized tissues, DAPS were subjected to unconfined compression testing to evaluate mechanical properties at ~0.5X body weight loading on the average human lumbar disc (0.24 MPa). CRISPR-engineered DAPS display significantly improved stress-strain curves compared to naïve hASCs with growth factors, indicating a stiffer tissue-engineered IVD (Fig. 8). Reductions in transition strain and maximum compressive strain were primarily responsible for driving the changes in mechanics observed in CRISPR-engineered DAPS compared to naïve hASCs dosed with growth factors. These changes are consistent with our histology and biochemical analysis demonstrating an increase in overall tissue deposition which increases the stiffness of the tissue. Increasing the maturation time period could further enhance DAPS mechanics, as progressive increases in mechanical properties have been shown previously with increasing culture duration [20].

This work has presented ZNF865 as an effective and useful tool to engineer stem cell phenotype utilizing CRISPR-guided gene modulation. Yet, this work does present limitations that will need to be addressed in future experiments. Initially, this work was solely completed in hASCs, a notoriously poor cartilage-producing cell type [67,68], and evaluating the effect of ZNF865-upregulation in additional cell types will need to be completed to better understand its applicability across cell types and disease states. We have investigated the therapeutic potential of ZNF865 in multiple cell types (primary human NP cells, HEK293, Jurkat) [45], however additional cell types will need to be evaluated as well. In addition, CRISPR-engineering a cell type that has a higher propensity for chondrogenic differentiation, such as bone marrow-derived mesenchymal stem cells, could further enhance matrix deposition compared to hASCs [67,68]. This work was solely completed in vitro, and in vivo testing will need to be completed to evaluate the efficacy of CRISPR-engineered DAPS and the inhibition of hypertrophy and mineralization. As has been previously shown, once implanted in vivo the harsh environment further limits functional tissue deposition, GAG and collagen retention, integration [20,24], and potential for hypertrophy and mineralization [66]. These edits are intended to overcome these challenges but will need to be evaluated in vivo in future work to better evaluate its potential as a cell and tissue engineering therapy.

This study investigates the use of targeted regulation of ZNF865 (BLST) as a tool in cell and tissue engineering, while also investigating the basic biology of this gene, which is largely unstudied. We demonstrate the basic biology of ZNF865 as a regulator of cell proliferation and protein processing, and as a cell engineering tool with potential application to a range of potential cell therapeutics. For example, we show that targeted upregulation of ZNF865 can be used to increase proliferation rates of cells for the rapid expansion of cell therapies [45]. With the recent FDA approval of a CRISPR-engineered cell therapy to treat sickle cell anemia [69,70], there is promise in engineering cell therapies to treat a host of diseases utilizing CRISPR. Furthermore, we present ZNF865 as a highly effective tool for enhancing cell therapy outcomes by significantly enhancing functional tissue deposition, manufacturing rates, and biomechanical properties of human-scale engineered tissues, while not resulting in chondrogenic hypertrophy. The proof-of-concept study in DAPS [20] displays the effectiveness of our ACAN/Col2a1-ZNF865-edited hASCs to produce functional fibrocartilage-like tissues. Overall, this work displays the efficacy and promise of CRISPR-guided gene modulation to engineer stem cells to develop more effective cell therapies to improve therapeutic outcomes of cell therapies and provide more effective treatments for musculoskeletal diseases.

Supplementary Material

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.actbio.2024.11.007.

Data availability

All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials.