Protocol to stress human primary pancreatic islets with diabetogenic factors for subsequent single-cell RNA sequencing

Summary

Type 1 diabetes (T1D) is an autoimmune disease that destroys insulin-secreting pancreatic β cells. Using human islets is instrumental to understanding the initiation and progression of T1D. Here, we present a protocol for treating cadaveric primary human islets with inflammatory, immune, and viral stressors. We detail Coxsackievirus B3 expansion, primary islet cell culture, and treatment with diabetogenic stressors, followed by procedures for single-cell RNA sequencing analysis.

For complete execution details, please refer to Maestas et al.¹ and Veronese-Paniagua et al.²

Graphical abstract

Highlights

• •Protocol for inducing viral, inflammatory, ER, and Golgi stress in human islets
• •Standardized workflow for multiplexed single-cell RNA sequencing of stressed islets
• •Enables discrimination of direct viral infection from bystander stress responses
• •Adaptable for both primary human islets and stem cell-derived islets

Publisher’s note: Undertaking any experimental protocol requires adherence to local institutional guidelines for laboratory safety and ethics.

Type 1 diabetes (T1D) is an autoimmune disease that destroys insulin-secreting pancreatic β cells. Using human islets is instrumental to understanding the initiation and progression of T1D. Here, we present a protocol for treating cadaveric primary human islets with inflammatory, immune, and viral stressors. We detail Coxsackievirus B3 expansion, primary islet cell culture, and treatment with diabetogenic stressors, followed by procedures for single-cell RNA sequencing analysis.

Before you begin

Type 1 diabetes (T1D) is an autoimmune disease marked by the destruction of insulin-secreting pancreatic β cells within the islets of Langerhans. Although genetic susceptibility is a major component of disease progression, environmental factors have also been postulated as triggers of disease pathology by inducing β cell stress. However, the heterogeneous composition of islets has limited our understanding of cell-specific responses to various forms of diabetogenic stress. This protocol describes the steps to expand Coxsackievirus B3 (CVB3) using existing viral stocks and HeLa cells. We then discuss the culture of human primary islets isolated from cadaveric donors and describe how to induce cellular stress in islets using various stressors, including CVB3 infection, and agents that induce endoplasmic reticulum and Golgi Apparatus stress, including Thapsigargin, Brefeldin A, and a variety of cytokines. We also detail the preparation of islets for single-cell RNA sequencing to delineate the cell-type-specific transcriptional responses to these stressors. Overall, this protocol will yield high-resolution, integrated transcriptomic datasets that will enable the identification of cell-state-specific signatures associated with islet dysfunction. By applying these methods, researchers can quantify viral transcript burden within individual cells to distinguish between direct infection and bystander effects. Furthermore, these datasets will allow for the identification of organelle-specific stress signatures, such as endoplasmic reticulum and Golgi Apparatus stress pathways induced by chemical stressors. These outputs facilitate downstream analyses, including differential gene expression across heterogenous cell populations and the discovery of novel signaling pathways involved in β cell death and destruction, serving as a robust foundation for identifying novel therapies that could delay or prevent T1D progression. Culture media should be prepared as described below before commencing this protocol.

Innovation

This protocol presents a refined methodological framework for dissecting the cell-type-specific transcriptional responses of primary human islets under diverse diabetogenic conditions, allowing for the identification of unique versus convergent stress signatures across heterogeneous cell populations. While traditional islet research often relies on bulk analysis, this method provides a standardized pipeline for multiplexing treatment conditions, including viral infection and organelle specific stress, which reduces batch effects and technical variability while maintaining donor-specific insights. Notably, the protocol enables the stratification of cells by viral load, distinguishing between direct infection and bystander effects. It is also uniquely adaptable, providing specific modifications for use with human pluripotent stem cell-derived islets, facilitating a direct comparison between primary tissue and emerging regenerative models. This comprehensive approach allows for the identification of specific genes and pathways involved in the initiation of type 1 diabetes, facilitating the discovery of novel therapeutic targets to prevent or delay β cell destruction.

Institutional permissions

Approval from the relevant ethics and safety committees is required to perform experiments using donor pancreatic islets or human pluripotent stem cells. Activities for this protocol were approved by the Washington University Institutional Biological & Chemical Safety Committee (Approval number 12189). All work involving HUES8 was approved by the Washington University Embryonic Stem Cell Research Oversight Committee (Approval number 15-002).

Preparation of primary islet culture medium

Timing: 2 min

Primary human islet culture medium consists of CMRL 1066 media supplemented with 10% Fetal Bovine Serum (FBS).

• 1.Add 50 mL FBS to 450 mL CMRL 1066 media.
• 2.Aliquot 40 mL of the medium in centrifuge tubes. The medium can be stored at 4^oC for 4 weeks.

Preparation of HeLa cell culture medium

Timing: 5 min

The HeLa cell culture medium consists of Dulbecco’s Modified Eagle Medium (DMEM) media supplemented with L-Glutamine, Sodium Pyruvate, Penicillin/Streptomycin, and heat inactivated fetal bovine serum.

• 3.Add 10 mL L-Glutamine (200 mM, final concentration 4 mM) to 430 mL high glucose Dulbecco’s Modified Eagle Medium (DMEM).
• 4.Add 5 mL Sodium Pyruvate (100 mM, final concentration 1 mM) to 430 mL DMEM.
• 5.Add 5 mL of pre-dissolved Penicillin/Streptomycin (10,000 units/mL, final concentration 100 units/mL) to 430 mL DMEM.
• 6.Add 50 mL heat-inactivated fetal bovine serum to 430 mL DMEM.

Note: HeLa cell culture medium can be stored at 4^oC for up to a month. Discard any unused media.

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
TotalSeq™ -A0251 anti-human Hashtag 1 Antibody (1 μg of antibody per 300 μL of Cell Staining Buffer)	BioLegend	394601; RRID: AB_2750015
TotalSeq™ -A0252 anti-human Hashtag 2 Antibody (1 μg of antibody per 300 μL of Cell Staining Buffer)	BioLegend	394603; RRID: AB_2750016
TotalSeq™ -A0253 anti-human hashtag 3 antibody (1 μg of antibody per 300 μL of cell staining buffer)	BioLegend	394605; RRID: AB_2750017
TotalSeq™ -A0254 anti-human Hashtag 4 Antibody (1 μg of antibody per 300 μL of Cell Staining Buffer)	BioLegend	394607; RRID: AB_2750018
TotalSeq™ -A0255 anti-human Hashtag 5 Antibody (1 μg of antibody per 300 μL of Cell Staining Buffer)	BioLegend	394609; RRID: AB_2750019
TotalSeq™ -A0256 anti-human Hashtag 6 Antibody (1 μg of antibody per 300 μL of Cell Staining Buffer)	BioLegend	394611; RRID: AB_2750020
Bacterial and virus strains
CVB3-eGFP/CVB3-Woodruff	Hubert Tse	N/A
Biological samples
Healthy primary human islets	Prodo Laboratories Inc.	N/A
Chemicals, peptides, and recombinant proteins
CMRL 1066	Corning	Cat#99-603-CV
DMEM, high glucose	ThermoFisher	Cat#11-965-092
L-Glutamine (200 mM)	ThermoFisher	Cat#25030-081
Sodium pyruvate (100 mM)	ThermoFisher	Cat#11360-070
Penicillin-Streptomycin (10,000 U/mL)	ThermoFisher	Cat#15140-122
Trypsin-EDTA (0.05%)	ThermoFisher	Cat#25300-054
TrypLE Express	Life Technologies	Cat#12604039
Y-27632	Pepro Tech	Cat#129382310MG
hESC-qualified Corning Matrigel Matrix	Corning	Cat#354277
Polyinosinic-polycytidylic acid [poly(I:C)]/LyoVec	InvivoGen	Cat#tlrl-piclv
Brefeldin A (BFA)	Sigma	Cat#B5936
Thapsigargin (Tg)	Sigma	Cat#T9033
Recombinant Human Interferon gamma (IFNγ)	R&D Systems	Cat#285F100
Recombinant human interleukin 1 beta (IL-1β)	R&D systems	Cat#201LB010/CF
Recombinant Human Tumor Necrosis Factor alpha (TNFα)	R&D Systems	Cat#210TA020/CF
1X PBS	Corning	Cat#21-030-CV
Fetal Bovine Serum (FBS)	ThermoFisher	Cat#A5256701
Heat-inactivated fetal bovine serum	Cytiva HyClone	Cat#SH3008803
Cell Staining Buffer	BioLegend	Cat#420201
Critical commercial assays
Chromium Single Cell 3′ V3.1 Library and Gel Bead Kit	10× Genomics	Cat#PN-1000128
Software and algorithms
Seurat V4.3.0.1	Hao, Y., et al.3 Cell. https://doi.org/10.1016/J.CELL.2021.04.048	https://satijalab.org/seurat/articles/install_v5
R V4.0.3	R	https://cran.rstudio.com/bin/windows/base/old/4.0.3/
R Studio V1.3.1093	RStudio	https://posit.co/download/rstudiodesktop/
Other
Countess 3 FL	ThermoFisher	N/A
NovaSeq6000	Illumina	N/A

Materials and equipment

Materials

HeLa cell culture medium

Reagent	Final concentration	Amount
DMEM, high glucose	N/A	430 mL
L-Glutamine (200 mM)	4 mM	10 mL
Sodium Pyruvate (100 mM)	1 mM	5 mL
Penicillin-Streptomycin (10,000 U/mL)	100 U/mL	5 mL
Heat-inactivated fetal bovine serum	10%	50 mL
Total	N/A	500 mL

Store at 4^oC for up to one month.

Primary islet culture medium

Reagent	Final concentration	Amount
CMRL 1066	N/A	450 mL
Fetal Bovine Serum	10%	50 mL
Total	N/A	500 mL

Store at 4^oC for up to one month.

Step-by-step method details

Initiating HeLa cell culture and infecting cells for CVB3 expansion

Timing: 6–8 days

This step describes how to start HeLa cell culture from cryopreserved stocks and expand CVB3 for downstream infection studies.

Note: All tissue culture procedures are performed in biological safety cabinets under sterile conditions, while wearing personal protective equipment that corresponds with biosafety level 2 (BSL2) regulations and virus work. All culture media and Trypsin-EDTA should be warmed to 37^oC prior to starting this step.

• 1.Using a 15 mL centrifuge tube, prepare a 9 mL aliquot of the HeLa cell culture medium made by following the “preparation of HeLa cell culture medium” section.
• 2.
  Thaw 1 vial containing 1 × 10⁶ total cryopreserved HeLa cells.

  • a.
    Take vial out of liquid nitrogen storage.
  • b.
    Immediately place vial in water bath set to 37^oC.
  • c.
    Remove vial immediately after ice crystals thaw.
• 3.Transfer HeLa cells into the 15 mL centrifuge tube detailed in step 1.
• 4.Centrifuge tube at 300 × g for 3 min.
• 5.Aspirate supernatant and resuspend the cell pellet in 10 mL of HeLa cell culture medium.
• 6.Transfer the resuspended cells into a T75 flask.
• 7.Transfer the flask into a humidified incubator set to 37^oC with 5% CO₂.
• 8.Allow the cells to expand for 2 to 3 days or until they reach 80-90% confluency.
• 9.Aspirate old culture medium from the flask and wash the cells with 10 mL sterile 1× PBS.
• 10.Aspirate the PBS and add 5 mL of Trypsin-EDTA.
• 11.Incubate the cells at 37^oC with 5% CO₂ for 5 min or until all cells are detached from flask.

Note: Single-cell suspension may take less than 5 min. Periodically check the cells by gently rocking the flask side-to-side or looking under a microscope.

• 12.Add 5 mL of fresh HeLa cell culture medium to the flask.
• 13.Transfer the cells to a 15 mL centrifuge tube.
• 14.Centrifuge the cells at 300 × g for 3 min.
• 15.Aspirate the supernatant and resuspend the cell pellet in 10 mL of HeLa cell culture medium.
• 16.Add 9 mL of fresh medium to each of two new T75 flasks.

Note: One flask will only be used for obtaining cell counts prior to inoculation with CVB3, while the other flask will actually be inoculated.

• 17.Seed 1 mL of the resuspended cells in each flask.
• 18.Transfer the flask into a humidified incubator set to 37^oC with 5% CO₂.
• 19.Allow the cells to expand for 2 to 3 days or until they are 80% confluent.
• 20.Once the cells are 80% confluent, follow steps 9–15 above using only one of the two T75 flasks.
• 21.Count the cells using either a hemocytometer or an automated cell counter and calculate the total cell number.
• 22.Discard the cells in the centrifuge tube.
• 23.Calculate the volume of virus needed to inoculate the cells in the second flask at a multiplicity of infection (MOI) of 0.1.

Note: To calculate the viral volume needed to infect with a MOI of 0.1, use the equation below where PFU (Plaque-Forming Units) represents the number of infectious virus particles and viral titer is the concentration of the virus (PFU/mL):

$$\left(0.1PFU\times Totalcellnumber\right)÷Viraltiter\left(\frac{PFU}{mL}\right)$$

• 24.Thaw a vial containing CVB3 corresponding to the viral titer used in the equation above.
• 25.Add the virus to 5 mL of HeLa cell culture medium.
• 26.Aspirate old culture medium from the remaining flask.
• 27.Add 10 mL of fresh HeLa cell culture medium.
• 28.Add 5 mL of virus-containing media.
• 29.Transfer the flask into an incubator set to 37^oC with 5% CO₂.
• 30.Allow infection to proceed for 36 h.

Note: After 36 h of infection, the cells should be floating, which indicates dead cells. This is the desired outcome since cell death is caused by the cytopathic effect of CVB3.

• 31.Transfer cells and culture medium to a 50 mL centrifuge tube.
• 32.Centrifuge the cells at 300 × g for 5 min.
• 33.Aspirate the supernatant and resuspend the cells in 1 mL of HeLa cell culture medium.
• 34.Transfer the cell solution to a cryo-vial.
• 35.Snap freeze cells by placing the cryo-vial in liquid nitrogen.

Note: This step is necessary to induce cell lysis and release the viral particles into the supernatant for later use. Alternatively, you can snap freeze the cells using dry ice in 100% ethanol.

• 36.Quickly transfer the cryo-vial into a 37^oC water bath until the media is thawed.
• 37.Repeat steps 35 and 36 two more times (total of three snap freezes)
• 38.Transfer the cell solution to a 15 mL centrifuge tube.
• 39.Centrifuge the tube at 300 × g for 5 min.
• 40.Transfer the supernatant, which contains the CVB3 virus stock, to a new tube.
• 41.Perform a plaque assay to measure the viral titer of the stock.

Note: Alternatively, store the supernatant at −80^oC until ready to perform the plaque assay.

Culture of human primary islets

Timing: 2 days

This step details the 3-dimensional culture of human primary islet clusters following delivery under controlled temperature conditions. The islets can then be used to model stress following treatment with various factors.

• 42.Pre-warm primary islet culture medium in a water or bead bath set to 37^oC.

Note: Pre-warm 6 mL of primary islet culture medium for every 2000 islet equivalent (IEQ). If working with less than 2000 IEQ and seeding in one well of a 6-well plate, then pre-warm 6 mL only.

Note: An IEQ is defined as an islet with a diameter of 150 μm.⁴ The number of IEQ in a Prodo Labs shipment is given in the order form sent with the islets.

• 43.Add 4 mL of islet culture medium to each corresponding well of a 6-well plate.
• 44.Transfer the islets from the delivery container to a 50 mL centrifuge tube using a 10 mL serological pipette.
• 45.Centrifuge the islets at 100 × g for 1 min.
• 46.Aspirate the medium and gently resuspend the islets in culture medium.

Note: Resuspend in 1 mL of culture medium for every well of a 6-well plate. For example, if seeding three wells of a 6-well plate, then resuspend the islets in 3 mL of islet culture medium.

• 47.Evenly distribute the islets across the corresponding wells by adding 1 mL of the islet solution per well.

Note: Gently resuspend periodically by mixing 2 times with a serological pipette to guarantee even distribution if seeding multiple wells. The total well volume should be 5 mL of culture medium. There should be 1000–2000 IEQ per well.

• 48.Transfer the plate to an Orbi-Shaker set at 100 RPM.

Note: The Orbi-Shaker should be placed in a humidified incubator at 5% CO₂ and 37^oC.

• 49.Culture the islets for at least 48 hours before starting any treatment.

CRITICAL: Feed islets every 48 h by performing a partial media change. Remove around 3 mL of culture media without removing any islets. To do this, swirl the plate until the islets are in the middle of the well. Tilt the culture dish and remove 3 mL of media from the edge of each well with a 5 mL serological pipette. Add 3 mL of fresh islet culture medium per well of a 6-well plate, bringing the total media volume to 5 mL.

Infection of human primary islets with CVB3

Timing: 1–2 days

This step details the acute infection of human primary islet clusters with CVB3. CVB3 is utilized to model the environmental hypothesis of T1D, where enteroviral infection is thought to trigger or accelerate autoimmune destruction by inducing direct cytopathic effects and inflammatory signaling.⁵ Notably, this protocol is designed to capture acute cellular responses rather than a chronic or persistent stress state. A defined chronic state is unfeasible due to the finite viability window of primary human islets in ex vivo 3D culture and the absence of an endogenous immune system to modulate the infection, which would otherwise lead to widespread, non-specific cell death over extended periods. Despite these temporal limitations, the generated results can be utilized to characterize the effects of acute viral infection on islet cell types. Steps 51-59 can also be followed if using human pluripotent stem cell-derived islets (SC-islets) with minor modifications to media composition and by performing a cell count following single-cell dispersal.⁶

• 50.Using a light microscope, count the number of islets per well.

Note: Perform at least two counts per well. Calculate the average of the counts.

• 51.To estimate the total number of cells per well, multiply the islet cluster number by 1,000.^7,8,9

Optional: Transfer a spare well of islets to a 15 mL centrifuge tube and follow steps 71–84 to obtain a more accurate total cell number.

If infecting SC-islets, transfer the cells to a 15 mL centrifuge tube and follow steps 71–84 to obtain cell count.

• 52.Calculate the viral volume needed to infect the islets at a MOI of 20.

Note: To calculate the viral volume needed to infect with a MOI of 20, use the following equation:

$$\left(20PFU\times Totalcellnumber\right)÷Viraltiter\left(\frac{PFU}{mL}\right)$$

• 53.Add 4.5 mL of fresh islet culture medium per well of a new 6-well plate.

Note: Add media to at least two wells.

• 54.Thaw a vial containing CVB3 at the corresponding viral titer.
• 55.Add the virus to at least one well of the 6-well plate from step 52.
• 56.Add an equal volume of endotoxin-free water or HeLa cell culture medium to another well of the 6-well plate from step 52.

Note: This well serves as the negative control.

• 57.Transfer islets by pipetting 500 μL of solution to each corresponding well.
• 58.Transfer the plate to the Orbi-Shaker in the incubator.
• 59.Culture the islets for the desired time point.

Treatment of human primary islets with diabetogenic stress compounds

Timing: 1–2 days

This step describes the treatment of 3D cultured human primary islets with various diabetogenic stress compounds to induce cellular stress. Brefeldin A (BFA) and Thapsigargin (Tg) were selected to model the intrinsic organelle stress signatures observed in failing islets.¹ Tg induces endoplasmic reticulum (ER) stress by depleting calcium stores, while BFA disrupts Golgi-mediated protein trafficking.¹ A cytokine mixture composed of Interferon gamma (IFNγ), Tumor Necrosis Factor alpha (TNFα), and Interleukin 1 beta (IL-1β) models the pro-inflammatory signaling and subsequent signaling cascades and ER stress that lead to immune-mediated β cell death.^10,11,12 The generated results can be utilized to characterize the effects of double-stranded RNA, as well as ER, inflammatory, or Golgi Apparatus (Golgi) stress on islet cell types using various functional and transcriptional assays. Steps 61–69 can also be followed if using SC-islets with minor modifications to drug concentrations and media composition.^1,5

• 60.
  Reconstitute poly(I:C), BFA, Tg, IFNγ, TNFα, and IL-1β according to manufacturers’ instructions.

  • a.
    Poly(I:C): https://www.invivogen.com/polyic-hmw-tlr3-agonist?mtm_campaign=ADS-LIGANDS-TLR3.
  • b.
    BFA: https://www.sigmaaldrich.com/US/en/product/sigma/b5936.
  • c.
    Tg: https://www.sigmaaldrich.com/US/en/product/sigma/t9033.
  • d.
    IFNγ: https://www.rndsystems.com/products/recombinant-human-ifn-gamma-protein_285-if.
  • e.
    TNFα: https://www.rndsystems.com/products/recombinant-human-tnf-alpha-protein_210-ta.
  • f.
    IL-1β: https://www.rndsystems.com/products/recombinant-human-il-1-beta-il-1f2-protein_201-lb.
• 61.
  Add the compounds to primary islet culture medium such that the final concentrations are the following:

  • a.
    Poly(I:C)/LyoVec = 500 ng/mL.
  • b.
    BFA = 1 μg/mL.
  • c.
    Tg = 10 μM.
  • d.
    IFNγ = 1000 ng/mL.
  • e.
    TNFα = 500 ng/mL.
  • f.
    IL-1β = 100 ng/mL.

Note: The three cytokines can be mixed in any permutation using the final concentrations provided above. A proper negative control, such as an equivalent volume of DMSO, should be used in these experiments as well. If using SC-islets, the final concentrations should be adjusted for BFA and IFNγ in the following way: BFA (0.1 μg/mL), IFNγ (500 ng/mL).

• 62.Add 4.5 mL of the corresponding condition to a well of a 6-well plate.
• 63.Transfer the primary islets to a 1.5 mL tube using a P1000 micropipette.
• 64.Allow the islets to settle for 2 min.
• 65.Remove the culture medium without disturbing the islets at the bottom of the tube.
• 66.Add 500 μL of the corresponding condition.
• 67.Transfer the primary islets to the corresponding well.
• 68.Transfer the plate to the incubated Orbi-Shaker.
• 69.Culture the islets for either 24 or 48 h.

Primary islet preparation for single-cell RNA sequencing

Timing: 3–5 h

This step details the preparation of 3D cultured human primary islets treated with various diabetogenic stress compounds or infected with CVB3 for single-cell RNA sequencing submission. Samples are multiplexed using BioLegend TotalSeq™ A antibodies for cell hashing, a method where each biological sample is labeled with a unique Hash Tag Oligo. This allows multiple “hashed samples” to be pooled and run in a single lane of a 10X Genomics Chromium chip, while maintaining the ability to demultiplex and identify the sample origin of each cell during downstream analysis. These pooled samples are then processed for library preparation and sequencing using the Chromium Single Cell 3′ v3.1 Library and Gel Bead Kit.

• 70.Keeping the conditions separate, transfer the treated islets to a 15 mL centrifuge tube.
• 71.Allow the islets to settle for 2 min.
• 72.Remove the culture medium without disturbing the islets at the bottom of the tube.
• 73.Perform a wash by adding 5 mL of 1× PBS.
• 74.Allow the islets to settle for 2 min.
• 75.Remove the PBS without disturbing the islets at the bottom of the tube.
• 76.Add 5 mL of TrypLE Express for 2000–4000 IEQ.
• 77.Transfer the tube to a 37^oC water bath.
• 78.Incubate the islets for 5 min.
• 79.Gently swirl the tube and place back in water bath.
• 80.Repeat steps 78 to 79 two more times.

Note: Total incubation time in TrypLE Express is 15 min.

• 81.Add 5 mL PBS to the centrifuge tube.
• 82.Centrifuge the tube at 300 × g for 3 min.
• 83.Aspirate the supernatant and add 1 mL of cold PBS.
• 84.Gently disperse the islets with a P1000 micropipette until the islets are single cell dispersed.
• 85.Wash the cells two more times with cold PBS for a total of three washes.
• 86.After the final wash, centrifuge the tube at 300 × g for 3 min.
• 87.Place the centrifuge tube on ice.
• 88.Prepare the TotalSeq™ A antibody cocktails by adding 1 μg of each antibody to 300 μL of Cell Staining Buffer.

Note: The number of antibodies used will correspond to the number of sample conditions that will be sequenced.

• 89.Centrifuge each antibody cocktail at 14,000 × g at 4^oC for 10 min.
• 90.Place the antibody cocktails on ice.

Note: Steps 88–90 should be done simultaneously with steps 81–87.

• 91.Aspirate the supernatant from the centrifuge tube without disturbing the cell pellet.
• 92.Resuspend each treatment condition in the corresponding TotalSeq™ A antibody cocktail.
• 93.Incubate the cells for 30 min at 4^oC.
• 94.Centrifuge the cells at 350 × g for 5 min at 4^oC.
• 95.Aspirate the supernatant without disturbing the cell pellet.
• 96.Resuspend the cells with 1 mL of cold Cell Staining Buffer using a p200 micropipette.
• 97.Centrifuge the cells at 350 × g for 5 min at 4^oC.
• 98.Repeat steps 95-97 two more times for a total of three washes.
• 99.Resuspend the cells with 1 mL cold DMEM.
• 100.Count the cells using a Countess 3 FL.
• 101.Centrifuge the cells at 350 × g for 5 min at 4^oC.
• 102.Aspirate the supernatant.
• 103.Resuspend the cells in cold DMEM to reach a cell concentration of 1,000 cells/μL.
• 104.Pool the samples at equal proportions.

Note: Do not pool CVB3-infected samples with non-infected conditions since there may be virions released into the solution from cell lysis.

• 105.Repeat steps 100–103.

Optional: Alternatively, you can prepare the cells for single-cell RNA sequencing using the 10× fixed RNA kit. Follow steps 70–82 of this protocol prior to commencing fixation and sample preparation according to the manufacturer's instructions.

https://www.10xgenomics.com/support/flex-gene-expression/documentation/steps/sample-prep/fixation-of-cells-and-nuclei-for-chromium-single-cell-gene-expression-flex.

https://www.10xgenomics.com/support/flex-gene-expression/documentation/steps/library-prep/chromium-single-cell-gene-expression-flex-reagent-kits-for-multiplexed-samples.

Note: Proceed to library preparation and construction using the 10X Chromium Single Cell 3′ v3.1 Library and Gel Bead Kit according to the manufacturer’s instructions.

https://www.10xgenomics.com/support/universal-three-prime-gene-expression/documentation/steps/library-prep/chromium-single-cell-3-reagent-kits-user-guide-v-3-1-chemistry.

Note: Upon completion of library construction, proceed to high-throughput sequencing. In this protocol, libraries were sequenced on an Illumina NovaSeq6000 System to generate the raw FASTQ files required for subsequent alignment, demultiplexing, and analysis.

Quality check and analysis of scRNA-seq data

Timing: 1–3 h

This step describes the raw data processing and filtering parameters applied to ensure only high quality single-cell data is retained for downstream analysis. We only use one dataset for this example, but this should be done with all hashed samples generated following the “primary islet preparation for single-cell RNA sequencing” section. Ensure your dataset is aligned to the human GRCh38 reference genome. If pCVB-eGFP was used, create a custom reference genome that includes the fusion protein sequence. Ensure your computer has at least 64-128 gigabytes of memory to successfully run the code in this section.

• 106.Download the datasets from CVB-infected experiments (GEO: GSE274264).
• 107.Create a folder and name it, “48hr-Control-sample_feature_bc_matrix”
• 108.
  Place the following files in the folder from step 107.

  • a.
    GSM8446415_patient1_48hr_control_barcodes.tsv.gz
  • b.
    GSM8446415_patient1_48hr_control_features.tsv.gz
  • c.
    GSM8446415_patient1_48hr_control_matrix.mtx.gz

CRITICAL: Remove GSM8446415_patient1_48hr_control_ from each of the files in step 108 (Ex. “GSM8446415_patient1_48hr_control_barcodes.tsv.gz” should be modified to “barcodes.tsv.gz”).

• 109.Download the Seurat package.³

> library(Seurat)

> library(dplyr)

> library(patchwork)

> library(ggplot2)

Note: This protocol uses R version 4.2.2, Matrix version 1.5–4.1, SeuratObject version 4.1.3, and Seurat V4.3.0.1 from https://satijalab.org/seurat/articles/install_v5

• 110.
  Implement quality check filtering to isolate high-quality single cells.

  • a.
    Load the raw featurebarcode matrix from the 48 hr control of patient 1 and apply stringent filtering criteria to remove technical noise.

#48hr Control Group-Patient01

> patient01.control48hr.data<-Read10X(“48hr-Control-sample_feature_bc_matrix”)

> patient01.control48hr.group<- CreateSeuratObject(counts = patient01.control48hr.data$'Gene Expression', min.cells=0, min.features = 20)

> patient01.control48hr.group[[“percent.mt”]] <-

PercentageFeatureSet(patient01.control48hr.group, pattern= “ˆMT-”)

> VlnPlot(patient01.control48hr.group, features = c(“nFeature_RNA”, “nCount_RNA”, “percent.mt”), ncol=3)

> plot1 <- FeatureScatter(patient01.control48hr.group, feature1 = “nFeature_RNA”, feature2 = “percent.mt”)

> plot2 <- FeatureScatter(patient01.control48hr.group, feature1 = “nCount_RNA”, feature2 = “nFeature_RNA”)

> plot1 + plot2

> patient01.control48hr.group <- subset(patient01.control48hr.group, subset= nFeature_RNA > 1000

& nFeature_RNA < 6000 & percent.mt < 15 & nCount_RNA < 30000)

> patient01.control48hr.group@meta.data$orig.ident<- “48hr Control”

> patient01.control48hr.group@meta.data$orig.patientID<- “Patient 01”

Note: Repeat these steps with each hashed sample for additional practice. Make sure to filter out barcodes with low gene complexity (<1000 features) to exclude empty droplets, and those with excessively high counts or features to mitigate doublet interference. Also, set a mitochondrial DNA (mtDNA) threshold (e.g., < 15%) to exclude apoptotic or stressed cells resulting from pre-sequencing handling. Figure 1A illustrates the corresponding data prior to quality check filtering. A pre-compiled R list object with all filtered samples is also provided in the next section.

Figure 1.

scRNA-seq analysis of stress-treated human primary human islets

(A) Violin plots showing the distribution of cells prior to quality check in patient 01 within the 48 hr control condition.

(B) UMAP depicting distribution of cells for each condition in the scRNA-seq dataset.

(C) UMAP of cell populations identified from scRNA-seq data integration.

(D) Feature plots depicting cell marker expression within clusters.

(E) UMAP showing original β cell population distribution (subset from all conditions).

(F) UMAP showing subpopulations after re-clustering β cells only (subset from all conditions).

(G) UMAP of β cell subpopulations grouped by viral load (subset from Control and CVB3 conditions).

Normalization, integration, clustering, cell-type identification, and transcriptional analysis

Timing: 1–3 days

The following step outlines how to utilize the quality-controlled data from the previous section to normalize, integrate, and cluster the scRNA-seq data prior to cell type identification. To streamline this process and avoid redundant quality check for every sample, we provide a pre-compiled R list object containing all filtered samples ready for downstream analysis. This section concludes with a detailed pipeline for identifying CVB3-infected cells, splitting cells by viral load, and performing differential gene expression analysis and in-depth cell-type-specific analyses.

• 111.Download the CVB3.list RDS file from GEO: GSE274264.

Note: This list contains all quality check filtered samples discussed in the section above.

CRITICAL: Load the file into R and ensure the object is renamed, “CVB3.list,” prior to starting the next step.

• 112.
  Perform batch correction and data integration to enable a unified analysis across all islet donors and stress conditions.

  • a.
    Normalize each dataset independently using SCTransform, regressing out mitochondrial content to mitigate the influence of mechanical cell stress on downstream clustering.
  • b.
    Following normalization, identify conserved biological anchors across samples to integrate the data into a single object (Figure 1B).

> library(sctransform)

> library(glmGamPoi)

#Perform normalization and feature selection using SCTransform

> CVB3.list <- lapply(X=CVB3.list, FUN = function(x) {

x <- SCTransform(x, method = “glmGamPoi”, vars.to.regress = “percent.mt”, verbose = FALSE)

})

#Select features that are repeatedly variable across datasets to use for downstream integration.

#Then prepare the object list normalized with SCTransform for integration

> CVB3.features <- SelectIntegrationFeatures(object.list= CVB3.list, nfeatures= 3000)

> CVB3.list <- PrepSCTIntegration(object.list = CVB3.list, anchor.features = CVB3.features)

#Perform integration by identifying anchors

> CVB3.anchors <- FindIntegrationAnchors(object.list= CVB3.list, anchor.features = CVB3.features, normalization.method = “SCT”)

> CVB3.integrated <- IntegrateData(anchorset = CVB3.anchors, normalization.method = “SCT”)

#Now you can run a single integrated analysis of all cells.

> CVB3.integrated <- RunPCA(CVB3.integrated, verbose=FALSE)

> CVB3.integrated <- RunUMAP(CVB3.integrated, dims = 1:30)

> CVB3.integrated <- FindNeighbors(CVB3.integrated, dims = 1:30)

> CVB3.integrated <- FindClusters(CVB3.integrated, resolution = 0.4)

#graphing by condition

> DimPlot(CVB3.integrated, reduction= “umap”, group.by = “orig.ident”, pt.size=0.5, shuffle = TRUE, cols = c(“#C6C0C3”, “#737072”,”#79D3E4”, “#5795A0”, “#BAA43D”, “#7D7130”))

Note: This integration step is critical for identifying shared cell-type clusters and performing robust comparative analyses between control, CVB3-infected, and chemically stressed islets while controlling for donor-specific batch effects.

• 113.Identify the differentially expressed genes for each cluster by performing a Wilcoxon Rank Sum test-comparing one cluster to all others.

Note: Utilize these cluster-specific marker genes in conjunction with known canonical islet markers to assign cell-type identities to each cluster (Figures 1C and 1D).

> DefaultAssay(CVB3.integrated) <- “SCT”

> CVB3.integrated<-PrepSCTFindMarkers(CVB3.integrated)

#Identifying the DEG in each cluster across clusters

> CVB3.integrated.DEG.markers <- FindAllMarkers(CVB3.integrated, only.pos = TRUE, min.pct = 0.25, logfc.threshold = 0.25)

#Identifying population IDs

> FeaturePlot(CVB3.integrated, features=c(“INS”, “DLK1”, “MAFA”, “UCN3”)) #β cell markers

> VlnPlot(CVB3.integrated, features = c(“INS”, “DLK1”, “MAFA”, “UCN3”), pt.size = 0)

> FeaturePlot(CVB3.integrated, features= c(“GCG”, “ARX”, “LOXL4”, “CRYBA2”)) #α cell markers

> VlnPlot(CVB3.integrated, features = c(“GCG”, “ARX”, “LOXL4”, “CRYBA2”), pt.size = 0)

> FeaturePlot(CVB3.integrated, features= c(“SST”, “LEPR”)) #δ cell markers

> VlnPlot(CVB3.integrated, features=c(“SST”, “LEPR”), pt.size=0)

> FeaturePlot(CVB3.integrated, features=c(“PPY”)) #γ cell markers

> VlnPlot(CVB3.integrated, features= c(“PPY”), pt.size=0)

> FeaturePlot(CVB3.integrated, features=c(“KRT17”, “KRT7”, “CFTR”, “SOX9”, “CRP”, “AQP1”)) #Ductal cell markers

> VlnPlot(CVB3.integrated, features = c(“KRT17”, “KRT7”, “CFTR”, “SOX9”, “CRP”, “AQP1”), pt.size = 0)

> FeaturePlot(CVB3.integrated, features=c(“PRSS1”, “SPINK1”, “KRT7”, “REG3A”, “CTRB2”)) #Acinar cell markers

> VlnPlot(CVB3.integrated, features= c(“PRSS1”, “SPINK1”, “KRT7”, “REG3A”, “CTRB2”), pt.size = 0)

> FeaturePlot(CVB3.integrated, features=c(“COL3A1”, “SPARC”)) #Mesenchyme markers

> VlnPlot(CVB3.integrated, features= c(“COL3A1”, “SPARC”))

> FeaturePlot(CVB3.integrated, features=c(“SPARC”, “ESAM”)) #Endothelial cell markers

> VlnPlot(CVB3.integrated, features = c(“SPARC”, “ESAM”))

> FeaturePlot(CVB3.integrated, features=c (“TMSB4X”, “CD14”, “FCGR3A”, “ITGAM”)) #Monocytes markers

> VlnPlot(CVB3.integrated, features=c(“TMSB4X”, “CD14”, “FCGR3A”, “ITGAM”))

• 114.Download the CVB3.integrated.markers.prepped RDS file from GEO: GSE274264.

CRITICAL: Load the file into R and ensure the object is renamed, “CVB3.integrated,” prior to starting the next step.

• 115.Rename and reorder the clusters with the identified cell types (Figure 1C).

#Renaming clusters

> new.cluster.ids <- c(“Alpha Cells 1”, “Beta Cells 1”, “Ductal Cells 1”, “Gamma Cells”, “Alpha Cells 2”, “Acinar Cells”, “Mesenchymal Cells”, “Delta Cells”, “Beta Cells 2”, “Adipocytes”, “Endothelial Cells”, “Monocytes”, “Alpha Cells 3”, “Ductal Cells 2”, “Mesenchymal Cells”, “Endothelial Cells”, “Mesenchymal Cells”)

> names(new.cluster.ids) <- levels(CVB3.integrated)

> CVB3.integrated <- RenameIdents(CVB3.integrated, new.cluster.ids)

> CVB3.integrated$cell_type <- CVB3.integrated@active.ident

#Reordering the clusters

> CVB3.integrated$cell_type <- factor(CVB3.integrated$cell_type, levels = c(“Alpha Cells 1”, “Alpha

Cells 2”, “Alpha Cells 3”, “Beta Cells 1”, “Beta Cells 2”, “Delta Cells”, “Gamma Cells”, “Ductal Cells 1”, “Ductal Cells 2”, “Acinar Cells”, “Mesenchymal Cells”, “Adipocytes”, “Endothelial Cells”, “Monocytes”))

> Idents(CVB3.integrated) <- “cell_type”

> DimPlot(CVB3.integrated, reduction= “umap”, pt.size=0.5, cols = c(“#2C82FF”, “#0049B3”,

“#A4E8FE”, “#AF37FC”, “#4D047E”, “#B4EC3A”, “#FBC600”, “#D78421”, “#905C47”, “#FF42D7”,

“#FF0000”, “#C4C4C4”, “#FF7E7E”, “#000000”))

• 116.
  Perform differential gene expression analysis to identify cell-specific stress signatures.

  • a.
    Conduct a pairwise comparison between CVB3-infected cells and their corresponding time matched controls utilizing the FindMarkers function.

##Subsetting the Beta cell populations

> Idents(CVB3.integrated) <- “cell_type”

> beta <- subset(CVB3.integrated, idents = c(“Beta Cells 1”, “Beta Cells 2”))

##Obtaining the statistically significant DEG for CVB conditions > Idents(beta) <- “orig.ident”

> CVB3.beta.24hr.markers <- FindMarkers(beta, ident.1=“24hrs CVB3”, ident.2 = “24hr Control”, assay = “SCT”, recorrect_umi=FALSE, only.pos = FALSE, min.pct = 0.25, logfc.threshold = 0)

> CVB3.beta.24hr.markers <- subset(CVB3.beta.24hr.markers, CVB3.beta.24hr.markers$p_val_adj < 0.05)

> CVB3.beta.48hr.markers <- FindMarkers(beta, ident.1 = “48hrs CVB3”, ident.2 = “48hr Control”, assay = “SCT”, recorrect_umi = FALSE, only.pos = FALSE, min.pct = 0.25, logfc.threshold = 0)

> CVB3.beta.48hr.markers <- subset(CVB3.beta.48hr.markers, CVB3.beta.48hr.markers$p_val_adj < 0.05)

Note: Subset each cell type and repeat the code above with the corresponding cell ID.

• 117.Download the CVB3.integrated.FINAL.rds RDS file from GEO: GSE274264.

CRITICAL: Load the file into R and ensure the object is renamed, “CVB3.integrated,” prior to starting the next step.

• 118.
  Identify CVB3-infected cells and quantify infection efficiency by detecting eGFP transcripts.

  • a.
    Run the following code to subset the CVB3-infected and control samples.
  • b.
    Categorize cells within the infected group as “GFP-Positive” (directly infected) or “GFP Negative (bystander).

Note: This categorization allows for the calculation of infection proportions across different islet cell types (Table 1).

CVB3.integrated <- readRDS(file=“/CVB3.integrated.FINAL.rds”)

#subsetting CVB3 and Control conditions

> Idents(CVB3.integrated) = “orig.ident”

> CVB3.GFP <- subset(CVB3.integrated, idents = c(“48hrs CVB3”, “48hr Control”))

#Identifying the cells with positive eGFP expression

> GFP.pos <- WhichCells(CVB3.GFP, idents = c(“48hrs CVB3”),expression = `pCVB-eGFP` > 0)

#Categorizing cells in CVB3.GFP as either GFP Positive, GFP Negative, or Control cells

> added_column = c()

> length = ncol(CVB3.GFP)

> count = 1

> cell.GFP.ID <- colnames(CVB3.GFP) %in% GFP.pos

> while (count <= length) { if(cell.GFP.ID[count]){

 value = “GFP Positive”

 }

 else if(CVB3.GFP$orig.ident[count] %in% c(“48hrs CVB3”)){ value = “GFP Negative”

 }

 else if(CVB3.GFP$orig.ident[count] %in% c(“48hr Control”)){ value = “Control”

 }

 > added_column = append(added_column, value)

 > count = count + 1

}

> View(added_column)

> CVB3.GFP$gfp.ident <- added_column

#Proportions of cells in each cluster--determining infection efficiency of CVB3

> Idents(CVB3.GFP) = “orig.ident”

> CVB.48h.GFP <- subset(CVB3.GFP, idents = c(“48hrs CVB3”))

> Idents(CVB.48h.GFP) = “gfp.ident”

> CVB.48h.GFPpos <- subset(CVB.48h.GFP, idents = c(“GFP Positive”))

> CVB.48h.GFPneg <- subset(CVB.48h.GFP, idents = c(“GFP Negative”))

> Idents(CVB.48h.GFPpos) = “cell_type”

> table(Idents(CVB.48h.GFPpos))

> Idents(CVB.48h.GFPneg) = “cell_type”

> table(Idents(CVB.48h.GFPneg))

• 119.
  Perform sub-clustering of β cells to achieve higher resolution for viral load analysis.

  • a.
    Prior to stratifying cells by viral load, isolate the β cell population and perform sub-clustering.

Note: This step recalculates the UMAP space and cluster neighbors specifically for the β cells, ensuring that the downstream differential expression analysis captures the subtle shifts in the β cell transcriptome caused by CVB3 infection.

#Reclustering beta cells

> Idents(CVB3.integrated) <- “cell_type”

> beta.subclust <- subset(CVB3.integrated, idents = c(“Beta Cells 1”, “Beta Cells 2”))

> DefaultAssay(beta.subclust) <- “integrated”

> beta.subclust <- RunUMAP(beta.subclust, dims = 1:30)

> beta.subclust <- FindNeighbors(beta.subclust, dims = 1:30)

> beta.subclust <- FindClusters(beta.subclust, resolution = 0.2, graph.name = NULL)

> Idents(beta.subclust) <- “seurat_clusters”

> DimPlot(beta.subclust, reduction = “umap”, label = FALSE, cols = c(“#BE51FF”, “#A26FB3”, “#5B267C”, “#E8A0FF”))

> Idents(beta.subclust) <- “cell_type”

> DimPlot(beta.subclust, reduction = “umap”, label = FALSE, cols = c(“#AF37FC”, “#4D047E”))

• 120.
  Stratify β cells by viral load using percentile-based eGFP expression thresholds.

  • a.
    Categorize the sub-clustered β cells into high, low, and bystander populations based on the distribution of eGFP transcripts.

Note: This stratification is achieved by calculating the 50^th percentile of viral transcript counts across the infected samples. This quantitative approach allows for the dissection of dosage-dependent transcriptional responses and the identification of pathways specifically activated by high levels of viral replication versus those triggered by the inflammatory environment (bystander effect).

> DefaultAssay(beta.subclust) <- “SCT”

> Idents(beta.subclust) <- “orig.ident”

> beta.subclust.GFP<- subset(beta.subclust, idents = c(“24hrs CVB3”, “48hrs CVB3”, “24hr Control”,

“48hr Control”))

> gene_of_interest <- “pCVB-eGFP”

> gene_expression_matrix <- GetAssayData(beta.subclust, assay = “RNA”)

# Calculate the percentile value

> percentile_threshold <- 50

> threshold_value <- quantile(gene_expression_matrix[gene_of_interest, ], probs = percentile_threshold / 100)

#Categorizing cells as either Bystander, Low GFP, or High GFP cells

> bystander <- WhichCells(beta.subclust.GFP, idents = c(“24hrs CVB3”, “48hrs CVB3”),expression = `pCVB-eGFP` == 0)

> GFP.low <- WhichCells(beta.subclust.GFP, idents = c(“24hrs CVB3”, “48hrs CVB3”), expression = `pCVB-eGFP` > 0 & `pCVB-eGFP` < threshold_value)

> GFP.high <- WhichCells(beta.subclust.GFP, idents = c(“24hrs CVB3”, “48hrs CVB3”), expression = `pCVB-eGFP` >= threshold_value)

> added_column = c()

> length = ncol(beta.subclust.GFP)

> count = 1

> cell.bystander.ID <- colnames(beta.subclust.GFP) %in% bystander

> cell.GFP.low.ID <- colnames(beta.subclust.GFP) %in% GFP.low

> cell.GFP.high.ID <- colnames(beta.subclust.GFP) %in% GFP.high

> while (count <= length) { if(cell.bystander.ID[count]){

 value = “Bystander”

 }

 else if (cell.GFP.low.ID[count]){

 value = “Low GFP”

 }

 else if (cell.GFP.high.ID[count]){

 value = “High GFP”

 }

 else if(beta.subclust.GFP$orig.ident[count] %in% c(“24hr Control”, “48hr Control”)){ value = “Control”

 }

 > added_column = append(added_column, value)

 > count = count + 1

}

> beta.subclust.GFP$gfp.ident2 <- added_column

> Idents(beta.subclust.GFP) = “gfp.ident2”

> DimPlot(beta.subclust.GFP, reduction = “umap”, shuffle = TRUE)

Table 1.

Table listing the number of CVB3-infected and bystander cells within the 48 h CVB3 condition

Cell type	CVB3-infected cells	Bystander cells
α Cells 1	1035	299
α Cells 2	448	114
α Cells 3	99	43
β Cells 1	785	312
β Cells 2	199	56
δ Cells	167	43
γ Cells	706	80
Ductal Cells 1	700	117
Ductal Cells 2	52	8
Acinar Cells	205	90
Mesenchyme	258	117
Adipocytes	127	13
Endothelial Cells	126	25
Monocytes	75	21

Expected outcomes

This protocol was developed to generate high-resolution, single-cell datasets documenting the transcriptomic shifts in primary human islets across diverse stress states. Researchers can expect to produce a multiplexed scRNA-seq dataset where individual cells are accurately assigned to their donor of origin and specific treatment condition. A key outcome is the ability to distinguish cell-type-specific vulnerabilities, such as differential gene expression signatures unique to β cells compared to α cells. For viral studies, the protocol specifically enables the quantification of viral transcript abundance within individual cells, allowing for the stratification of data into high-load, low-load, and bystander populations. This granular data facilitates the identification of both direct viral-induced damage and indirect signaling effects. Ultimately, the analysis outputs provide a map of islet dysfunction, serving as a robust foundation for identifying and validating novel therapeutic targets intended to preserve β cell mass and function.

Limitations

A fundamental limitation of this protocol is the donor-to-donor variability inherent to primary human islets. To perform robust experiments, perform the outlined steps with at least a sample size of three patients. Larger sample sizes derived from patients from different ethnic groups, ages, sexes, and other metrics will also capture inter-donor variability and the broader patient population heterogeneity. Alternatively, using SC-islets is another viable option. Additionally, this protocol does not establish a persistent stress state, which may better model long-term complications. This protocol uses relatively high MOI for CVB3 infection. Although 3D-cultured cells require higher treatment concentrations than their 2D counterparts,¹³ and our treatments are consistent with previous studies,^{2,14,15,16,17,18} these conditions may lead to a lack of cell specificity. Computationally, a limitation of this protocol is that the computational workflow was validated using only specific combinations of software versions. Different software versions may necessitate modifications, particularly in terms of how they interact with each other. Additionally, integration of large scRNA-seq datasets may require substantial memory resources depending on system configuration.

Troubleshooting

Problem 1

Low HeLa cell viability or slow confluency prior to CVB3 expansion (Step 20).

Potential solution

Delayed transfer or overheating during the thawing process could decrease cell viability. Ensure the cryovial is removed from the 37^oC water bath immediately after ice crystals disappear to prevent DMSO toxicity.

Also, verify that the DMEM high-glucose medium is properly supplemented with the correct final concentrations of L-Glutamine, Sodium Pyruvate, Penicillin/Streptomycin, and heat-inactivated FBS.

Problem 2

Absence of cytopathic effect or low viral titer after 36 hours of CVB3 infection of HeLa cells (Step 30).

Potential solution

Re-verify the total cell number of the counting flask prior to inoculation and strictly follow the MOI equation to calculate the volume of virus to add to the cells:

$$\left(0.1PFU\times Totalcellnumber\right)÷Viraltiter\left(\frac{PFU}{mL}\right)$$

Also, ensure that the virus-infected cell pellet undergoes exactly three complete freeze-thaw cycles to completely lyse the cells and release viral particles into the supernatant before performing the plaque assay.

Problem 3

High primary islet mortality or disintegration during the initial 48-hour culture (Step 49).

Potential solution

Excessive physical stress or severe shear forces during pipetting and handling could induce islet death. Handle primary human islets with extreme care, avoiding vigorous mixing. Maintain a strict density of 1,000-2,000 IEQ per well of a 6-well plate to avoid overcrowding or significant nutrient depletion during culture. Ensure the culture plate is placed on an orbital shaker set continuously at 100 RPM inside the humidified incubator to guarantee constant nutrient diffusion and aggregate suspension.

Problem 4

Primary islets are inadvertently lost or aspirated during scheduled medium replenishment (Step 49).

Potential solution

Failure to centralize the islet aggregates before aspirating can lead to cluster lost during aspiration. Always swirl the 6-well plate gently in a circular motion until the islet clusters gather completetly in the center of the well. Tilt the culture plate carefully and slowly aspirate exactly 3 mL of the spent medium from the outer edge of the well using a serological pipette.

Problem 5

Primary islets do not undergo a strong response to stress treatments (Steps 59 and 69).

Potential solution

Repeat the treatment with additional patients to account for donor heterogeneity.

Problem 6

Lack of or limited access to patient material (Section: “culture of human primary islets”).

Potential solution

SC-islets can also be used following proper approvals by the relevant Institutional Biological & Chemical Safety Committee and Embryonic Stem Cell Research Oversight Committee.

Problem 7

Incomplete single-cell dissociation of islets leading to cell loss or microfluidic clogging (Steps 70–84).

Potential solution

Ensure the TrypLE Express is warmed to 37^oC prior to using it on the cells. The incubation time can be extended past the 15-minute mark but should not exceed 18 minutes. Ensure gentle swirling of the tube every 5 minutes. After adding cold PBS, use a P1000 micropipette to pipette the suspension up and down meticulously and gently until no visible islet fragments or cell clumps remain.

Problem 8

Poor sample demultiplexing or cross-contamination of cell hashes in downstream analysis (Steps 88–105).

Potential solution

To avoid insufficient washing of unbound TotalSeq™ A antibodies, adhere strictly to the three distinct post-staining washes utilizing cold Cell Staining Buffer to remove residual, unbound antibody conjugates.

Never pool CVB3-infected cells with non-infected or chemically stressed conditions prior to microfluidic processing, as viral particles released from lysed cells can cross-infect negative control populations.

Problem 9

R script crashes with an “Out of Memory” error during data integration or clustering (Step 112).

Potential solution

Ensure the computational workstation features at least 64–128 gigabytes of physical RAM to execute the SCTransform normalization and anchor-based data integration steps across multiple donors. If system memory is limited, subset datasets to lower cell counts or downsample non-essential populations before running FindIntegrationAnchors.

Problem 10

R script crashes with an error indicating that a function or argument has been deprecated or removed.

Potential solution

Explicitly verify and lock the active package version in your R session using packageVersion(“Seurat”). Major version updates often deprecate or restructure core functions. If a newer version is active, install and load the specific validated archive version using the remotes package to prevent syntax mismatches:

> library(remotes)

> remotes::install_version(“Seurat”, version = “4.3.0.1”)

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Jeffrey R. Millman (jmillman@wustl.edu).

Technical contact

Technical questions on executing this protocol should be directed to and will be answered by the technical contact, Jeffrey R. Millman (jmillman@wustl.edu) or Hubert M. Tse (htse@kumc.edu).

Materials availability

This study did not generate new materials for this protocol.

Data and code availability

Sequencing datasets have been deposited into Gene Expression Omnibus (GEO: GSE237448 and GEO: GSE274264). Additional information may be found in Maestas et al.¹ and Veronese-Paniagua et al.²

Acknowledgments

This work was primarily funded by the National Institutes of Health (NIH) (R01DK138469) to J.R.M. and H.M.T., with additional support from NIH (R01DK114233, R01DK127497, and R01DK126456); Breakthrough T1D (3-SRA-2023-1295-S-B); the Edward J. Mallinckrodt Foundation; and Washington University School of Medicine Department of Medicine startup funds to J.R.M. D.A.V.-P. was supported by the National Science Foundation’s (NSF) Graduate Research Fellowship Program (DGE 2139839). Further support came from the Washington University Diabetes Research Center (P30DK020579). We thank the Genome Technology Access Center at the McDonnell Genome Institute at Washington University School of Medicine for genomic analysis support. The center is partially supported by NCI Cancer Center Support Grant P30CA91842 to the Siteman Cancer Center from the National Center for Research Resources (NCRR), a component of the NIH, and the NIH Roadmap for Medical Research. This publication is solely the authors’ responsibility and does not represent the official views of NIH, NSF, or any other funder. We would also like to thank Erika Brown (Washington University) for helpful feedback on the manuscript.

Author contributions

D.A.V.-P., M.M.M., and J.R.M. designed all the experiments and wrote the manuscript. D.A.V.-P. and M.M.M. generated all the data. J.P.T. and H.M.T. provided key guidance, insights, and reagents. All authors revised, reviewed, and approved the manuscript.

Declaration of interests

J.R.M. is an inventor on licensed patents and patent applications related to SC-islets. J.R.M. was employed at and has stock in Sana Biotechnology.