A protocol for metabolic characterization of human induced pluripotent stem cell-derived cardiomyocytes (iPS-CM)

Graphical abstract


Description of protocol
The discovery of cell reprogramming methods to obtain human induced pluripotent stem cells (iPSc) from adult somatic cells, followed by differentiation to cardiomyocytes, provides a unique opportunity to generate an unlimited number of induced pluripotent derives cardiomyocytes (iPS-CM) carrying specific cardiac genotypes and phenotypes. This invaluable model is an innovative tool for creating in vitro cardiac cellular models that can be utilized for new therapeutic assessments, screening of drug's induced cardiac toxic effects and a discovery of as-yet unknown metabolic mechanisms that underlie cardiac manifestation of the inherited heart diseases. The ability to successfully translate cell-based studies into a clinical studies requires efficient tools for functional characterization of iPSC-CM. For this purpose electrophysiological and imaging techniques are utilized. These methods allow to monitor ion channel activities [1], measure action potentials, changes in Ca +2 fluxes [2], mitochondria viability and apoptosis [3]. Here we report new metabolic assays that allow to expand functional analysis beyond already established techniques and provide metabolic characteristics of iPSCM for the purpose of disease modeling, drug candidates screening and therapeutic agents cardiac safety assessments.

Method details
Differentiation iPSC cells were kindly contributed by Duke University core facility. Differentiation to cardiomyocytes was induced by the modified small molecules protocol [4] (Fig. 1). Briefly hiPSCs (passage >20) were passed at a 1:12 ratio and cultured on matrigel with Essential 8 Flex medium (Thermo Fisher Scientific) until~100% confluency. At Day 0, prior to medium change, E8 media was aspirated from cells, and they were washed with D-PBS. 1 mL of 0.5 mM EDTA in DPBS was added to each culture well. Plates were placed in incubator for approximately 35 s. Upon removal from incubator, dishes were gently tapped to dislodge only a small amount of cells. EDTA solution was aspirated, and differentiation medium was added directly to the EDTA-treated cells. The medium was changed to RPMI 1640 with B27 minus insulin supplement (A1895601, Life Technologies) with 6 mM CHIR99021,10 ng/mL Activin A, and 50 ug/mL ascorbic acid. On day 1, medium was changed to RPMI 1640 with B27 minus insulin supplement (referred to as basal differentiation supplement). On day 2, medium was changed to basal differentiation supplement and 10 mM IWR-1. On day 4, medium was refreshed with basal supplement and 10 mM IWR-1. Medium was changed on Day 6 to basal differentiation supplement (without IWR-1). On Day 8, and every other day thereafter, medium was refreshed with RPMI 1640 with B27 supplement (referred to as cardiomyocyte maintenance media). Spontaneously contracting cardiomyocytes (Fig. 2, Supplemental video) were first observed on day 9. Approximately 2 days after observing initial cardiomyocyte contractions, medium was changed to RPMI 1640 with no glucose and supplemented with 4 mM lactate. Medium was changed 72 h later to cardiomyocyte maintenance media.

iPSCM harvest and extraction
All protocols below are adjusted for 1 well of a 6-well culture plate.

Materials
Labeled Metabolism quenching protocol 1. Aspirate culture media from iPSCM. 2. Wash cells twice with 1 mL cold PBS and one time with water. Ensure that water remains on cells for no more than 30 s. 3. Add 1 mL cold acetonitrile and incubate cells in À20 C for 20 min. At the end of the 20 min period white protein precipitate will appear on culture dish bottom. 4. Add 0.75 mL of cold water and with the plastic scraper, scrape cells from the bottom of the culture dish. 5. Using a P1000 pipette, remove cell lysate to a 10 mL centrifuge tube. 6. Repeat steps 3-5 one more time. To ensures complete recovery of lysate it is recommended to use same cells scraper and pipette tip. 7. Add internal standard. Internal standards preparation and amounts varies with the analysis type.
See notes #1-3. 8. To the collected cell lysate (step#4), add 1 mL of cold chloroform. 9. Vortex sample and centrifuge at 5000 rpm for 15 min. After completion of the centrifugation step, a clear separation will be observed between the polar (upper) phase, non-polar (lower) phase, and the inter-phase pellet that contains proteins and DNA. 10. From the upper polar phase, take an aliquot of 250 mL for acylcarnitines and amino acids assay and transfer to the clean tube. Label " amino acids and acylcarnitines". 11. Carefully remove and combine the rest of the upper polar phase with the bottom non-polar phase in a clean tube. Label "small cellular metabolites by GC/MS".
12. Add 50 mL methanol to the remaining protein pellet and evaporate solvent under nitrogen stream in room temperature. 13. Reconstitute protein with buffer and perform protein analysis by protein assay such as Bradford or bicinchoninic acid (BCA).
Add 100 mL of the working solution to the cell lysate.
Note#3-For the GCMS metabolic profile, add 25 mL of 1 mM tricarballylic acid 3. Cool to the room temperature and dry again.
4. Reconstitute with 100 mL of mobile phase A and transfer to vials for the analysis.

HPLC parameters
Introduce derivatized samples directly by the injection to the mass spectrometer instrument through the inline filter with no chromatographic separation. Keep inline filter in HPLC oven at 32 C.
Inject 7 mL of a sample into a flowing solvent at 50 mL/min flow rate for the total 1.8 run time. Use only mobile phase A isocratically.

MS/MS-tandem mass spectrometer
Use following tandem mass spectrometry scans for the high throughput: for the acylcarnitines precursor ion scan at m/z 85 (Table 1, only saturated C2-18 acylcarnitines are included), for amino acids neutral loss (NL 102, NL 119), and single reaction monitoring SRM m/z 231.1-m/z 70.1 for arginine ( Table 2).

Data analysis
Perform data analysis by Chemoview 2.2 software (SCIEX) or comparable software. Calculate peak areas at 50% peak height.     Data analysis. Mass spectra analyzed by freely available AMDIS software and Fiehn library (Agilent). Identified metabolites are shown in Fig. 5. Normalize metabolite's levels to the reference standard tricarballylic acid m/z 377 (as BSTFA derivative) as following: use mass spectra to calculate the ratio of peak areas of target metabolites to the reference internal standard (tricarballylic acid) and to the total protein amount of each cell pellet as determined by the Bradford assay.
Relative level ¼ Peak area of metabolite= Peak area of reference standard ð Þ ½ Ã reference standard in nmol ½ units Protein amount

Representative data
Glucose uptake, lactate and pyruvate production analysis Glucose and lactate represent a major carbon sources for the myocardial energy metabolism. Whereas pyruvate is produced from lactate through the reaction catalyzed by lactate dehydrogenase, lactate to pyruvate ratio is a good surrogate for cytosolic redox state and correlates well with NAD + /NADH ratio.
For glucose uptake and lactate/pyruvate production assays incubate cells with 13 C labeled glucose. We recommend to use a fully labeled commercial available 13 C 6 glucose. 13 C labeled glucose concentration in the cell media can be ranged from low (5 mM) to high (25 mM) concentration. Glucose uptake is measured by monitoring decrease in 13 C 6 Glucose concentration in culture media over the time (Fig. 8) and lactate and pyruvate production by the appearance of 13 C 3 labeled lactate (Fig. 8) and 13 C 3 Materials 13 C 6 labeled glucose (Sigma-Aldrich, cat.no 389374) 1 mM galactose (Sigma-Aldrich, cat.no PHR1206) Tricarballylic acid (Sigma-Aldrich cat.no T53503) prepare 1 mM in water Acetic anhydride (Sigma-Aldrich, cat.no 242845-5G) 4. Cool to the room temperature and transfer to GCMS vials.

GCMS parameters
Following ions were monitored: for 13 C 3 lactate m/z 264, for 13 C 3 pyruvate m/z 177 and m/z 377 for tricarballylic acid.

Data analysis
Note: Data analysis was performed under validated assumption that there is no significant natural abundance contribution to the 13 C 3 -Lactate and 13 C 3 pyruvate, thus for the concentration calculations the data was not corrected for 13 C natural abundance.
Relative level ¼ ½Peak area of 13 C 3 l lactate=ðPeak area of referencestandardÞÃ reference standard in ½nmol units Protein amount Relative level ¼ ½Peak area of 13 C 3 pyruvate=ðPeak area of reference standardÞÃ reference standard in ½nmol units Protein amount Representative chromatogram See Fig. 7.

Results and discussion
In recent years there are emerging evidence for studies involved human iPSCM in basic science, regenerative medicine, and the pharmaceutical industry. Although much progress has been made in the genetic and electrophysiological characterization of iPSCMs, it is not understood well how  metabolic phenotypes are affected in the course of the disease or in response to newly developed therapies. Thus, to fulfill the potential for iPSCM related studies, there is a need for metabolic assays complementary to the existing functional characterization methods. Here, we developed mass spectrometry-based quantitative assays and applied them for metabolic characterization of iPS-CM. Using these assays we successfully profiled a variety of small metabolites involved in different biochemical pathways.
Acylcarnitines have a central role in the transport of fatty acids into the mitochondria for subsequent β-oxidation. Acylcarnitine analysis can be especially useful for iPSCM models of heart failure [5,6], diabetic cardiomyopathy [7] and fatty acid oxidation disorders (FAOD) with cardiac manifestation [8,9]. Given the fact that the relative acylcarnitines pools and individual contributions within the total carnitine pool are closely correlated to the intra-mitochondrial acyl-CoA pools it is suggested that proposed assay is a powerful tool for the mitochondrial dysfunction investigations in iPSCM. Amino acids also have a high significance to heart metabolism. Although under physiological condition, the human heart has minimal reliance on amino acids for ATP, some amino acids replenish citric acid cycle and their utilization increases during heart failure, hypoxia or another pathological metabolic remodeling. The proposed high throughput assay analyses acylcarnitines and amino acids in two minutes and has an automated data analysis algorithm (Chemoview/ABSCIEX) but can be also performed manually.

GCMS metabolic profiling
The proposed gas chromatography-mass spectrometry metabolic profiling allows semiquantitative analysis of small metabolites (50-650 Da). Although method requires extensive sample preparation including chemical derivatization [10] to increase metabolites' volatility for gas chromatography, it was able to detect thirty two metabolites extracted from one well (standard six well plates, 1.2-1.5 * 10 6 iPS-CM per well in average). For metabolites identification we applied freely available deconvolution software AMDIS [11] and commercially available spectral library [12]. Under standardized EI-GCMS conditions, every metabolite produced unique fragmentation pattern that was matched to the library based on >78% match criteria. The identified small metabolites included amino acids, carbohydrates, fatty acids, and sterols.

Lactate/pyruvate assay
Pyruvate is the end product of glycolysis and it also can be produced from lactate through the reaction catalyzed by lactate dehydrogenase. Thus, lactate to pyruvate ratio is a good surrogate for cytosolic redox state and correlates well with NAD+/NADH ratio. Dysfunctions in a citric acid cycle or respiratory chain may lead to reduced pyruvate oxidation and to the abnormal lactate to pyruvate ratios. At the same time, excessive lactate production is an indication of lactic acidosis which is detrimental to the iPSCM contractility.

Glucose uptake assay
In iPSCM cellular system, glucose is a primary energy source, thus uptake assay is an important tool. Alterations in glucose uptake can reflect overall metabolic changes, and in conjunction with the pyruvate production can indicate glycolysis flux changes. At the same time, glucose uptake monitoring can be indicative of the over or down expression of glucose transporters.

Conclusion
We have developed a workflow for the analysis of small metabolites in iPSCM involved in energyrelated pathways. The workflow combines GCMS and LC-MS/MS platforms serves as an important tool for iPSCM metabolic characterization and can provide complementary metabolic endpoints to monitor therapeutic interventions. Future studies to expand iPSCM metabolic profiles are warranted.