Genetically Engineered Macrophages Derived from iPSCs for Self-Regulating Delivery of Anti-Inflammatory Biologic Drugs

,


Introduction
Rheumatoid arthritis (RA) is a painful and debilitating joint disease afecting 1% of the population worldwide [1].Te pathogenesis of RA is believed to be mediated by increased levels of proinfammatory cytokines, such as interleukin (IL)-1α/β and tumor necrosis factor alpha (TNF-α), which drive chronic infammation, cartilage destruction, and pain [2][3][4][5].Advances in disease-modifying antirheumatic drugs (DMARDs) have led to the development of therapies that can reduce joint infammation and pain, two of the leading causes of work disability in RA patients.However, while partially efective, these treatments are often delivered at high, immunosuppressive doses, which can predispose patients to signifcant and debilitating of-target efects while also failing to fully mitigate disease [6,7].Furthermore, current treatments do not provide the temporal specifcity and precise control necessary to treat fuctuating RA symptoms.In this regard, a self-regulating cell-based therapy could provide a sensitive, safe, and specifc system for the long-term delivery of biologic drugs in a therapeutic manner, mitigating adverse events triggered by conventional RA biologics [8].
A candidate cell type for cell-based biologic drug delivery is the macrophage, a key immune cell highly prevalent in RA that shows capabilities for homing and engraftment to sites of infammation [9].During chronic infammation, the dysregulated signaling by macrophages drives an imbalance of infammatory cytokines, accelerating the progression of disease [2,[10][11][12][13].Initial infammatory signaling by macrophages perpetuates the infammatory environment of RA through numerous downstream mediators including the recruitment of additional immune cells, induction of T-cell diferentiation, and increased bone resorption processes [11].Tus, targeting downstream byproducts of this infammation can limit harmful side efects.However, the selfpropagation (e.g., positive feedback loop) of infammatory signaling by macrophages makes attenuation of chronic infammation challenging.Furthermore, modifying cells of the innate immune system is an ongoing challenge, and novel approaches are necessary to harness macrophage signaling for RA treatment [14].Here, we propose the use of "smart" macrophages that possess the capacity to deliver precisely controlled anticytokine drugs in a self-regulating manner.
Te use of cell therapies based on pluripotent stem cells has gained signifcant interest in the feld of immunoengineering [15][16][17][18].Recent advances in the felds of synthetic biology and genome editing have enabled rapid and precise engineering of the cellular genome, allowing researchers to immunomodulate previously difcult targets for a wide array of applications [19][20][21][22][23]. CRISPR-Cas9 genome editing was used to reprogram murine-induced pluripotent stem cells (miPSCs) as a basis for developing cell-based biologic delivery.Specifcally, monocyte chemoattractant protein-1 (MCP-1, gene name Ccl2), a chemotaxis-inducing molecule produced downstream of TNF-α or IL-1 signaling, was targeted to create genetic circuits with a gene addition on one allele, as not to compromise overall cellular function, and to generate miPSCs encoding: (1) Ccl2-luciferase (Luc)-a frefy luciferase transcriptional reporter, or (2) Ccl2-sTNFR1-chimeric human sTNFR1murine immunoglobulin G (Figure 1(a)) [15].In these circuits, TNF-α signaling leads to the activation of NF-κB infammatory cascade and the induction of Ccl2 (Figure 1(b)).Subsequent release of sTNFR1 under the Ccl2 locus (Figure 1(c)) results in the inhibition of TNF-α signaling, ceasing induction of Ccl2 expression and sTNFR1 production (Figure 1(d)).Tus, these two genetic circuits act to confer cytokine-activated and feedback-controlled gene expressions, with the Ccl2-Luc line driving luciferase expression to allow imaging for circuit activation.In this study, we demonstrated the successful diferentiation and development of a novel iPSC-derived engineered macrophage (iMAC) capable of responding to infammation and delivering precisely controlled anticytokine drugs in a selfregulating manner.Tese engineered macrophages provide an efective proof of concept for an innovative framework extending previous systems of immunoengineered drug delivery vehicles for anti-infammatory applications.
L929 conditioned media was generated by culturing L929 cells for 7 days in DMEM/F12, 10% FBS, and 1% penicillin/streptomycin. Media were collected, fltered, and stored at −20 °C until use.Bone-marrow-derived macrophages were generated by frst isolating the bone marrow from long bones in C57BL/6 mice.Bone marrow was then incubated with red cell lysis bufer and strained with a 40 μm strainer.Isolated cells were cultured for 7-10 days in DMEM-HG with 10% FBS, 1% penicillin/streptomycin, and 30% L929 conditioned media prior to testing.

Flow Cytometry.
Cells were passed through a 40 μm strainer to remove debris and blocked with Fc receptor antibody CD16/32 to prevent nonspecifc antibody binding before staining.Dead cells were stained with propidium iodide and examined for diferentiation markers (Supplementary Table 1) in a cell staining bufer (Biolegend).Doublets and cellular debris were excluded through FlowJo analysis.For cell sorting, similar methods were utilized to stain CD45-APC positive cells, which were sorted using a FACS Aria II (BD Bioscience).

RNA Isolation.
Following experimental treatment, culture media were collected, and cells were imaged and then rinsed in PBS, lysed in RL bufer, snap frozen, and stored at −80 °C until RNA isolation.RNA was isolated according to the manufacturer's recommendations (Total RNA Purifcation Plus Kit; Norgen Biotek).

2
Journal of Tissue Engineering and Regenerative Medicine 2.4.Gene Expression with RT-qPCR.200 ng of RNA was reverse transcribed using SuperScript VILO complementary DNA master mix (Invitrogen).Real-time PCR was performed using Fast SyBR Green Master Mix (Applied Biosystems) on a QuantStudio (TermoFisher) with 10 ng of cDNA and primer concentration at 10 μM (Supplementary Table 2).All primers were validated prior to use between 90 and 110% efciency.All reactions were performed in duplicate for each analyzed gene.Diferences in gene expression were calculated as fold change using the ΔΔCT method: fold change was normalized to no treatment or day 0 with GAPDH as the housekeeping gene where GAPDH maintained consistent expression throughout all diferentiation timepoints and experimental treatments.
2.6.Boyden Cell Migration Assay.Cells were prepared by trypsinization and incubated in starvation media (media with 5% FBS) for 1 hr.Cells were resuspended in DMEM at a concentration of 5 × 10 5 cells/mL.Polycarbonate PFB flters (Neuro Probe) with 8 μm pores were used.Culture media with 20 ng/ml of TNF-α or 100 ng/ml MCP1 were placed in the bottom chamber of Neuro Probe 48-well chemotaxis chambers (Neuro Probe) with 5 × 10 4 cells in the top chamber and incubated at 37 °C in 95% O 2 /5% CO 2 for 24 h.Migrated cells were fxed in methanol and stained with 500 nM DAPI.

Additional Assays.
Luciferase activity from Ccl2-Luc cells in all monolayer experiments was measured using the Bright-Glo luminescence assay (Promega) and a Cyta-tion5 plate reader.Luciferase activity is reported as relative luminescence.
Culture media were collected from samples and stored at −20 °C.sTNFR1 concentration was measured with DuoSet enzyme-linked immunosorbent assay specifc to human sTNFR1 (R&D Systems).Each sample was measured in technical duplicates, and absorbance was measured at 450 and 540 nm.
As nitric oxide production has been well documented as a major player in infammation, nitric oxide production in the supernatant was measured using a Griess assay.Cell supernatant was collected, and the concentration of nitric oxide within the culture medium was measured by the Griess reagent (Promega) according to the manufacturer's instructions.
A phagocytosis assay (Cayman Chemical) was performed according to the manufacturer's instructions to measure the phagocytic ability of iMACs, BMDMs, and miPSCs.Briefy, cells were treated with FITC-labeled latex beads or control media for 4 hours (1 : 200).Excess beads were removed by washing with PBS, and cells were scraped in 1 mL of assay bufer, fxed, stained with CD14 and DAPI, and quantifed through fow cytometry for cellular uptake (BD X-20).
2.9.Statistical Analysis.Statistical analysis was performed with GraphPad Prism using analysis of variance (ANOVA) (α � 0.05) with Tukey's HSD post hoc test.For qRT-PCR comparisons, all data were normalized to no treatment or time zero as the control and fold-change values were logtransformed prior to statistical analysis.For all data, standard error of means was used (SEM).

Murine iPSCs Successfully Diferentiate into a Macrophage
Lineage.Te primary goal of this work was to investigate the potential of using CRISPR-Cas9 reprogrammed iPSCs to derive macrophages as an engineered cell therapy with the capacity to respond to an infammatory stimulus and deliver anticytokine drugs in a precisely controlled and selfregulating manner.Terefore, it was frst necessary to develop a protocol that could reliably diferentiate macrophages from miPSCs.Published protocols were tested to obtain an optimized diferentiation protocol for the successful development of macrophages from multiple wild type and edited miPSC lines [27][28][29][30].Briefy, this protocol relies upon embryoid body cell culture with daily feeding of a distinct set of growth factors and small molecules (Figures 2(a) and 2(b)).To characterize the generated iMACs, successive qPCR testing for diferentiation markers was completed at day 0, 4, 8, 10, and 17 over multiple diferentiations in two unique cell lines (Figure 2c; Supplementary Figure 1).As a point of reference, all timepoints were compared to day 0 of diferentiation.Gene expression throughout the protocol demonstrated distinct waves in expression.Early-stage pluripotency genes (Nanog and Oct4) were signifcantly downregulated by terminal diferentiation (Figure 2c).Key hemangioblast and hematopoietic genes (Klf4, Gata2, Pdgf-a, αSMA, and Flk1) increased in expression during primitive hematopoiesis (day 0-10) but lost expression by the terminal macrophage endpoint while macrophage diferentiation markers (Cd45 and Cd11b) increased over time.Additional fow cytometry and immunocytochemistry analysis were performed examining stem cell and macrophage markers.Flow cytometry for both early (CD117) and late (CD41 and CD45) markers of primitive hematopoiesis as well as myeloid precursor (CD34) further corroborated successful diferentiation (Figures 2(d) and 2(e), Supplementary Figure 1A).Immunocytochemistry on fully diferentiated iMACs showed macrophage marker expression (CD11b, CD14, and CD45) throughout the entire population by day 17 in comparison to an unstained control, while earlier lineage marker CD34 was not present and was confrmed by fow cytometry where diferentiated macrophages were largely CD45/CD11b/CD14 + and CD34 − (Figure 2(f ); Supplementary Figure 2).

iMACs Respond Similarly in Phenotype and Function to
Bone Marrow-Derived Macrophages.We then further characterized the phenotype of our diferentiated unedited miPSC-derived macrophages as functionally diferentiated macrophages as compared to the well-studied murine bone-marrow-derived macrophage (BMDM) model.Generated iMACs at day 17 and fully diferentiated BMDMs were stimulated for 24 hours in either no treatment, 100 ng/mL IFNc and 1 ng/mL LPS, or 10 ng/mL IL-4 and 10 ng/mL IL-13 and evaluated in terms of morphology, activation, and function.Cell morphology remained consistent between iMACs and BMDMs with treatment (Figure 3(a)).When stimulated with IFNc/LPS or IL-4/ IL-13 for 24 hours, qPCR gene expression analysis demonstrated that iMACs activated similarly to BMDMs, signifcantly increasing expression (p < 0.05, n � 4) of infammatory genes (Socs, Il6, Tnfa, Cd80, Stat1, Cd86, and Vegf ) in response to IFNc/LPS and immunomodulatory genes (Cd11c, Ym1, Cd206, Irf4, Il10, Cd163, Fizz, and Arg1) in response to IL-4/IL-13, in comparison to a nontreated control (Figure 3(b), Supplementary Figure 3).Interestingly, though similar in pattern, iMACs exhibited a dampened response to cytokine stimulus as compared to BMDMs, demonstrating an overall lower fold change in response to all cytokine treatments than BMDMs.Western blot analysis of NOS2, CD206, and arginase with a cyclophilin A/GAPDH loading control indicated similar regulation in protein expression by both iMACs and BMDMs with increased expression of NOS2 and arginase in response to IFNc/LPS and IL-4/IL-13 stimuli, respectively, with no response without treatment (Figure 3(c)).CD206 was expressed in all cells at a basal level but decreased in response to an infammatory stimulus and increased in response to an immunomodulatory stimulus.
In regard to function, both iMACs and BMDMs produced a signifcant amount of nitric oxide (p < 0.05, n = 3) in response to infammatory activation, though production by BMDMs was 2-fold higher (Figure 3(d)).Following incubation with latex beads, cells at the miPSC stage did not uptake beads as expected; both iMACs and BMDMs phagocytized beads with similar efciency of 50-60% cellular uptake after excluding CD14 -cells (Figure 3(e), Supplementary Figure 1B).To evaluate the ability of iMACs to home in response to both cytokine and chemokine signals, both iMACs and BMDMs were treated with either no treatment or IFNc/LPS for 24 hours to represent homing in response to infammatory signals.Following a 24-hour incubation in a Boyden chamber, BMDMs showed higher migration in response to both a cytokine (TNF-α) and chemokine (CCL2/MCP-1) gradient in comparison to iMACs, where low-level migration was only observed in response to CCL2/MCP-1 (Figure 3(f )).

Terapeutic iMAC Mitigated Infammatory
Signaling by the Self-Regulatory Production of sTNFR1.First, we evaluated whether iMACs expressing the sTNFR1 gene edit could produce a therapeutic in response to varying concentrations of an infammatory stimulus.iMACs were treated with a low and high dose of TNF-α (5 and 20 ng/mL), and mRNA and culture media were collected at 24 and 72 hours (Figures 4(a) and 4(b)).sTNFR1 and Il6 gene expressions were evaluated by qRT-PCR.At both TNF-α concentrations, sTNFR1 and Il6 gene expressions were increased compared with cells given no treatment (Figure 4(b)).However, in the 5 ng/mL group, there was signifcantly less infammatory activation observed and no change in expression between the 24-and 72-hour timepoints, while in the 20 ng/mL group, gene expression was signifcantly higher at the 24-hour timepoint than 72 hours, demonstrating downregulation in expression over time.To fully characterize the response of iMACs possessing either the luciferase (Ccl2-Luc) or Ccl2-sTNFR1 circuits to an infammatory stimulus, Ccl2-Luc/sTNFR1 iMACs were treated with 0 or 20 ng/ml of TNF-α for 72 hours.Ccl2-Luc iMACs treated with 20 ng/ml of TNF-α expressed signifcant production of luminescence after 24 hours that resolved to baseline by 72 hours after stimulation as compared to the untreated control (Figure 4(c)).Similarly, in response to stimulation by TNF-α (20 ng/ml), Ccl2-sTNFR1 iMACs produced signifcantly more protein at 24 hours than no treatment (Figure 4(d)).As expected, since luminescence was attenuated by 72 hours in Ccl2-Luc cells, there was also no diference in sTNFR1 production between no treatment and TNF-α treatment in Ccl2-sTNFR1 iMACs at 72 hours.Examining Ccl2 and sTNFR1 expression in Ccl2-sTNFR1 iMACs at both the 24-and 72-hour timepoints, expression normalized to an untreated control and confrmed that circuit activity was halted by 72 hours, demonstrating self-regulation in these two systems (Figure 4(e)).When Ccl2-sTNFR1 iMACs were examined for the repeatability of response to multiple fares, autoregulation of the circuit was maintained, where again sTNFR1 presence peaked by 24 hours following each stimulation with TNF-α and returned to baseline by 48 hours (Figure 4(f)).Tis production of sTNFR1 was able to mitigate infammatory activation in Ccl2-sTNFR1 iMACs as compared to control Ccl2-Luc iMACs, with peak protection against infammatory gene activation demonstrated at 24 hours and signifcant downregulation still maintained by 48 hours following stimulation (Figure 4(g)).Tis modulation was not exclusively TNF-specifc, with Ccl2-sTNFR1 iMACs also demonstrating protection in response to stimulation with IFNc/LPS and further immunomodulatory upregulation in response to IL-4/IL-13 that led to the decrease of secreted nitric oxide in response to infammatory activation (Figure 4(h)).However, the phagocytic capacity of Ccl2-sTNFR1 iMACs remained unchanged as compared to control Ccl2-Luc iMACs (Figure 4i).

Discussion
Here, we demonstrated a proof of concept for engineered iPSC-derived macrophages, iMACs, capable of sensing and dynamically responding to infammation by producing the 6 Journal of Tissue Engineering and Regenerative Medicine anti-infammatory mediator sTNFR1 in a self-regulated manner.Our results optimized the generation of iPSCderived macrophages, validating and extending previously published protocols.Macrophages diferentiated from miPSC showed successful sequential diferentiation into a hematopoietic stem cell-like and then macrophage lineage as confrmed by positive expression (or loss) of major differentiation factors characteristic of each stage of development.While iMACs responded to stimuli in a similar pattern to BMDMs, they displayed a lower response to both cytokine and chemokine activation that led to not only reduced gene activation but also reduced nitric oxide production and migration in comparison to BMDMs.Likewise, other studies have investigated the phenotypical diferences between stem cell and bone-marrow-derived macrophages that suggest a similarity between tissueresident macrophages and iPSC-and ESC-derived macrophages formed from the primitive and transient defnitive wave of hematopoiesis [31][32][33][34].Notably, as this diferentiation process relies upon the selective sorting of CD45 + cells, scaling up into larger-scale applications would require discussion of alternate culture strategies.As such, these fndings indicated a successful proof of concept for the generation of macrophages from miPSCs that resembled tissue-resident macrophages, and further modifcations of such protocols are needed to optimize macrophage diferentiation, polarization, and behavior.Ccl2 has been shown to be a crucial driver of the recruitment and proliferation of immune cells in the synovium, which results in the upregulation of infammation and cytotoxic molecules and disruption of the immune network in the joint [35,36].Terefore, Ccl2 was selected as the initial driver for therapeutic delivery.In response to TNF-α stimulation, iMACs possessing both the Ccl2-Luc and Ccl2-sTNFR1 gene circuits responded in a self-regulatory manner and observed through the activation and subsequent deactivation of the Ccl2 promoter that regulated luminescence and sTNFR1 protein production, respectively.Importantly, this regulation was both rapid and uniform, consistently peaking by 24 hours and resolving by 48-72 hours in response to both single and iterative stimulations.Notably, while transgene production was resolved by 48 hours, robust anti-infammatory efects persisted well beyond circuit autoregulation, with Ccl2-sTNFR1 iMACs maintaining signifcant downregulation of infammatory genes, and increased immunomodulatory genes as compared to Ccl2-Luc iMACs.While Ccl2-Luc iMACs did not exhibit any transgene expression basally, sTNFR1 protein secretion was detected in the absence of cytokine in sTNFR1 iMACs as macrophages have been known to produce low levels of immunomodulatory factors without the activation of cytokines.
Te emerging feld of synthetic immunology has highlighted immune cells as ideal targets for therapeutic reprogramming where these cells can be systematically engineered to detect a multitude of environmental inputs to initiate complex, nuanced, and controlled therapeutic responses [21,22,37].Macrophages provide an important target for cell therapies, as research continues to highlight their role in a broad range of diseases.Tey are strongly regulated by local environmental cues and possess an intrinsic homing ability, allowing them to migrate in response to these cues to sites of high infammation, like the arthritic joint [2,13,38].Indeed, the therapeutic potential of macrophage-based drug delivery strategies has been encouraging where macrophages have been used to effectively deliver multiple types of cargo [39][40][41].Our work extends these eforts through the generation and engineering of macrophages that are not only capable of delivering a protein drug cargo but also possess the ability to produce multiple drugs and autoregulate their production and delivery through preprogrammed synthetic gene circuits.
Overall, this study demonstrated the successful generation of engineered iMACs for the controlled, autoregulated exogenous induction of biologics.Broadly, we examined the use of engineered iMACs in response to TNF-α-mediated infammation; however, other transgenes can be utilized for developing autoregulated biological systems for a wide range of other infammatory stimuli and disease applications.Indeed, it is our hope that this approach will emphasize the strength of macrophages, in particular iMACs, as an important overlooked therapeutic target for disease therapy, and inspire future studies examining the development of more complex circuits.

Figure 1 :Figure 2 :
Figure 1: Depiction of the reprogrammed infammatory signaling pathway in CRISPR-Cas9-engineered iMACs for the biologic Ccl2-sTNFR1 (1) and reporter Ccl2-luciferase (2) circuits.(a) TNF signaling through the TNFR type 1 receptor initiates a cascade leading to nuclear translocation and increased transcriptional activity of NF-κB, activating an infammatory transcriptional program.(b) Te Ccl2 promotor is then activated, which induces the expression of soluble TNF type 1 receptor (sTNFR1) in the biologic circuit (1) and luciferase in the reporter circuit (2).(c) Upon antagonism of TNF in the microenvironment, signal transduction through TNFR1 and the NF-κB infammatory cascade is inhibited.(d) Expression of the sTNFR1 transgene decreases upon inhibition of NF-κB initiating infammation.

Figure 3 :
Figure 3: Polarization and signaling in response to cytokine treatment with either no treatment (NT), IFNc/LPS, or IL-4/IL-13 stimulus after 24 hours.(a) Phase contrast imaging of iMAC in comparison to BMDMs.(b) qPCR normalized to GAPDH demonstrated similar upregulated gene expression in response to infammatory or immunomodulatory stimuli, with lower overall fold-change response for iMACs than in BMDMs.Heat map color represents downregulation (blue), upregulation (red), or no change (grey) per gene (n � 6).(c) Western blot analysis of iMACs/BMDMs cytokine activation.(d) Griess assay for nitric oxide production of iMAC/BMDMs in response to infammatory stimuli.(e) Phagocytosis in iMACs/BMDMs quantifed by fow cytometry.(f ) Migration images from Boyden chamber for iMACs/BMDMs, naïve, or prestimulated with IFNc/LPS, in response to chemoattractants (n � 4).(d, e) Bars represent mean ± SEM with # (iMAC) or * (BMDM) denoting p < 0.05 to no treatment by one-way ANOVA.