A simple method for analyzing the differential gene expression of coronary endothelial cells and cardiac muscle cells was developed. Cells were isolated from guinea pig hearts by collagenase digestion. In the diluted cell suspension, single cardiomyocytes and capillary fragments containing 6–15 endothelial cells could be identified morphologically. A simple “cell picker” was constructed using a polyethylene pipette with a tip diameter of ∼150 μm that was attached to a micromanipulator and connected to an electric miniature valve. Intermittent suction pulses (1- to 2-cm water column) were applied by opening the valve for 100–200 ms at 1-s intervals. Cardiomyocytes (800–1,000) or capillary fragments (150) were picked under visual control using an inverted microscope. The cells were transferred to a reaction tube for RNA extraction, reverse transcription (RT), and DNA amplification (RT-PCR) with gene-specific and intron-spanning primers. All PCR products were verified by sequencing. Troponin T and endothelin-1 were found to be specific markers for guinea-pig cardiac muscle cells and coronary endothelial cells, respectively.
- troponin T
- capillary endothelium
- cell picker
one of the most exciting questions of cell biology is the regulation of cell-specific and organ-specific gene expression. To investigate differential gene expression, usually RNA is extracted from different organs and analyzed by RT-PCR or by Northern blot hybridization. However, every organ contains several cell types. In the mammalian heart, for example, only 30% of the cells are cardiac muscle cells (5), with the remainder consisting of endothelial cells, vascular smooth muscle cells, pericytes and fibroblasts, and nerve cells. It would be highly desirable to get more information on the cell-specific expression profile of the different cell types.
Single-cell RT-PCR has proved useful, especially in the field of neuronal ion channels and receptors, and it is still a considerable challenge to try and correlate the mRNA signals found in individual neurons with the ion channels and receptors in these cells (3). Correlation of the properties of heterologously expressed channels and receptors with the properties of channels present in native cells is difficult, because the properties of channels and receptors may change depending on the expression system used. Unfortunately, single cell RT-PCR is not feasible in all cell types. In endothelial cells, for example, the amount of cytoplasm that can be harvested with a patch pipette is extremely small, because the thickness of the cell outside the nuclear region is <1 μm. Moreover, gigaseal formation is impeded by the basement membrane surrounding the endothelial cell layer. Therefore, we have devised an alternative method to get information on the gene expression of coronary endothelial cells. The method is based on separation of different cell types with collagenase and visual recognition and selection of cells that can be identified by their morphology. The cells are taken up by a simple “cell picker,” transferred to a reaction tube, and then subjected to RT-PCR. As a test of the reliability and efficiency of the method, we compared the expression of endothelin-1 (ET-1) and troponin T in capillary endothelial cells and cardiomyocytes isolated from guinea pig heart.
Isolated cardiomyocytes and capillaries were obtained from guinea pig hearts by enzymatic dispersion. Guinea pigs weighing 250–450 g were decapitated and their hearts rapidly excised. A cannula was attached to the aorta (Langendorff preparation), and the coronary arteries were perfused at a constant flow rate of 10 ml/min using a peristaltic pump. The heart was submerged in a small organ bath warmed to 37°C. The perfusing solution contained (in mM) 130 NaCl, 15 KCl, 2 CaCl2, 0.8 MgCl2, 1 NaH2PO4, 2 Na-pyruvate, 10 glucose, and 10 HEPES. The pH was 7.4 (adjusted with NaOH); the temperature was 37°C. The elevated K+ concentration of the perfusate caused cardiac arrest. Within 15–20 min, coronary perfusion pressure increased to a steady level between 60 and 100 mmHg, indicating recovery of energy metabolism.
To initiate dissociation of the cells, the heart was perfused for 5 min with nominally Ca2+-free solution, which otherwise had the same composition as described above. The heart was then perfused for 10 min with Ca2+-free solution to which 30 μM Ca2+ and 1 or 1.5 mg/ml collagenase (Worthington, type II) were added. Subsequently, the heart was removed from the organ bath, washed briefly in a “storage solution” containing (in mM) 65 K-glutamate, 45 KCl, 30 KH2PO4, 3 MgSO4, 0.5 EGTA, 20 taurine, and 10 glucose (pH adjusted to 7.4 with KOH), and submerged in 30 ml of storage solution. The heart was then cut into small pieces, and the cells were suspended by gentle trituration with a Pasteur pipette.
Selection of cardiomyocytes.
Two milliliters of the mixed cell suspension were layered on storage solution containing 4% bovine serum albumin (BSA), and the heart cells were allowed to settle by gravity sedimentation for 15–20 min (4). This procedure was then repeated, and the final pellet was resuspended in 3–5 ml of storage solution. For cell picking, several drops of the enriched cardiomyocyte suspension were transferred to a 35-mm petri dish filled with storage solution to which 1% BSA had been added to prevent attachment of the cells. The petri dish with the diluted cell suspension was mounted on an inverted microscope (Zeiss IM 35). Up to 1,000 cardiomyocytes were collected under visual control with the cell picker.
Selection of capillary endothelial cells.
Single isolated endothelial cells round up and cannot be easily discriminated from fibroblasts, damaged vascular smooth muscle cells, or pericytes. Therefore, capillary fragments were used that can be easily recognized morphologically as a regularly spaced string of nuclei connected by a tube of 3–5 μm in diameter (2). The mixed cell suspension obtained as described above was sieved through a nylon filter (Falcon, mesh size 70 μm) that retained the capillary fragments. The cells were washed with physiological salt solution (PSS) containing (in mM) 135 NaCl, 5 KCl, 1 CaCl2, 1 MgCl2, 0.33 NaH2PO4, 10 glucose, 2 Na-pyruvate, and 10 HEPES (pH adjusted to 7.4 with NaOH). The capillary fragments were then removed from the sieve by rinsing with PSS. Single drops of the enriched endothelial cell suspension were added to a 35-mm petri dish containing PSS with 1% BSA. Roughly 150 capillary fragments consisting of 6–15 endothelial cells were collected under visual control with the cell picker. Care was taken to exclude capillary fragments to which cells were attached (putative pericytes) or that were contaminated with cell debris.
RNA extraction, RT, and PCR.
For the isolation of the total RNA, a commercial kit was used (RNeasy Mini Kit, Qiagen, Chatsworth, CA). Usually, 800–1,000 picked cardiomyocytes or roughly 150 capillary fragments were transferred to 800 μl of lysis solution (RNeasy Mini), and the total RNA was extracted according to the instructions of the manufacturer. One-third of the RNA eluate was reverse transcribed using 200 units of Superscript II reverse transcriptase (GIBCO BRL) and a random hexanucleotide mixture (total volume 25 μl). The PCR was performed with gene-specific, intron-spanning primers of 24–28 nucleotides in length. The total PCR volume was 50 μl, including 2 μl of RT reaction, 50 pmol of each primer, and 2.5 units of AmpliTaq Gold (Applied Biosystems). Thus one RT reaction gave ∼10 PCR runs. The tubes were placed in the thermal cycler (model 2400, Applied Biosystems), and the PCR was run with a hot start for 5 min at 94°C (initial melt); then for 40 cycles of 0.5 min at 94°C, 0.5 min at 55°C, and 1 min at 72°C; and, at last, for 5 min at 72°C (final extension). Fifteen microliters of the PCR probe were size fractionated by agarose gel electrophoresis. To verify the identity of the PCR products, DNA fragments of the expected length were isolated and directly sequenced using an ABI prism 310 DNA sequencer (Applied Biosystems).
To control the purity of the cell fractions, we designed intron-spanning primers for cell-specific markers. Troponin T and ET-1 transcripts were used as markers for cardiomyocytes and endothelial cells, respectively. Primers for glyceraldehyde-3-phosphate dehydrogenase (GAPDH), a constitutively and ubiquitously expressed gene, were used to check RT-PCR conditions. Specific primers for guinea pig troponin T were required to obtain distinct PCR signals. The sequences of the guinea pig-specific markers were submitted to GenBank. In the case of GAPDH, primers derived from the human sequence worked well.
Primers for guinea pig troponin T were as follows: sense, 5′-GACCTGAACGAGCTGCAGACGCTGATC-3′; and antisense, 5′-CTTTCTGGTTGTCATTGATCCTGTTCC-3′ (located on exons 8 and 15 in the rat gene, respectively). These primers amplified a DNA product of 530 bp in length (accession number: AF099071).
Primers for guinea pig ET-1 were as follows: sense, 5′-GTGGTCTCTGGAGCGGAGCTCAGC-3′; and antisense, 5′-CTTGGCAGAAATTCCAGCACTTCTTG-3′ (located on exons 2 and 3 in the mouse gene, respectively). These primers amplified a DNA product of 314 bp in length (accession number: S82654).
Primers for human GAPDH were as follows: sense, 5′-CATCACCATCTTCCAGGAGCGA-3′; and antisense, 5′-GTCTTCTGGGTGGCAGTGATGG-3′ (located on exons 3/4 and 7 in the human gene, respectively). These primers amplified a DNA product of 332 bp in length (accession number for the 295-bp guinea pig DNA fragment: AF099070).
To study cell-specific gene expression in the different cardiac cells, we have devised a method for the selective isolation of myocytes and capillary fragments. The principle of the design of the cell picker is illustrated in Fig. 1. It consisted of a polyethylene pipette with a tip diameter of ∼150 μm, an electric two-way microvalve, and a piece of silicon tubing serving as a siphon. The degree of suction applied could be adjusted by varying the level of the outlet of the tubing, which was usually 1–2 cm below the level of the petri dish mounted on the inverted microscope. The electric valve was a normally closed two-way valve, the opening of which was controlled by a pulse generator (+5 V). Voltage pulses of 20–200 ms in duration (controlled by a potentiometer) could be applied at regular intervals by pressing a knob. The standard interval chosen was 1 s, which was well above the reaction time of the experimenter. Thus the number of suction pulses could be adjusted to the minimum required to transfer the selected cell into the pipette.
The pipette was attached to a micromanipulator fixed to the table of the inverted microscope. The pipette was moved into the center of the field of vision 5–10 μm above the bottom of the petri dish. The microscope was equipped with a sliding plate that could be moved manually, independently of the microscope table to which the micromanipulator was attached. Cardiomyocytes and capillary fragments were enriched as described in methods(Fig. 2, Aand B). Cells showing no attached debris were selected visually at high magnification (×320) and moved below the opening of the pipette. Suction pulses were then applied until the cell was taken up into the pipette (Fig. 2,C andD). Subsequently, the sliding plate was moved to place the next cell under the opening of the pipette. With practice, 200 cardiomyocytes or 20 capillary fragments could be picked in 10 min.
The purity of the cell fractions was checked with gene-specific intron-spanning primers. Figure 3,A andB, shows that similar levels of GAPDH transcripts were found in both cell fractions, indicating that comparable amounts of total RNA were used in the experiments. Expression of the troponin T gene was detected only in cardiomyocytes (n = 15; Fig.3 A); no transcripts were found in capillary fragments (n = 6; Fig.3 B). The sequence of the amplified troponin T fragment revealed the cardiac muscle-specific isoform. ET-1 is normally produced by endothelium (6) and might therefore represent a specific marker for endothelial cells. ET-1 transcripts were found only in freshly isolated capillary endothelial cells (n = 6; Fig.3 B) and were never detected in freshly isolated pure ventricular cardiomyocytes of the guinea pig (n = 15; Fig.3 A).
We have shown that it is possible to isolate pure heart cell fractions with an easy and highly efficient method. The purity of the collected cell fractions was controlled by RT-PCR with gene-specific, intron-spanning primers. The cell picker described here requires only an inverted microscope, a micromanipulator, and, in addition, some plastic materials and an electric microvalve. Recently, other methods to select single cells or groups of cells have been tried, for example, laser-assisted microdissection (1). This approach is well suited for stained tissue sections, especially under pathophysiological conditions. However, for studying cell-specific gene expression, the method described here is probably simpler, more sensitive, and less expensive.
ET-1, a peptide that was originally isolated from aortic endothelium (6, 8), is one of the endothelium-derived contracting factors. We have detected expression of the ET-1 gene only in freshly isolated coronary endothelial cells and not in cardiomyocytes of guinea pig heart (Fig. 3). In contrast, Suzuki et al. (7) recently reported that ET-1 is expressed in cultured neonatal rat cardiac myocytes. This discrepancy may be attributable to the difference in species or age. However, it cannot be excluded that expression of ET-1 in neonatal rat cardiomyocytes was induced by the cell culture conditions (7). It is well known that the morphology and the pattern of gene expression changes considerably during cell culture. The difference between the results of Suzuki et al. (7) and our results underscores the importance of using freshly isolated cells for studying cell-specific gene expression.
The results presented here suggest that, at least in the adult guinea pig, ET-1 is not constitutively expressed in cardiomyocytes. ET-1 was found to be a specific marker for freshly isolated coronary capillary endothelial cells from guinea pig heart. Troponin T was highly expressed in cardiomyocytes and could not be detected in any of the endothelial cell preparations (Fig. 3). Thus troponin T appears to be a specific marker for cardiac muscle cells.
We have shown that multicell RT-PCR is a suitable method for the detection of gene transcripts in different cell types of the heart. mRNA transcribed at a very low rate can be detected, and, on the other hand, the expression of certain genes in specific cell types can be reliably excluded. Furthermore, the expression of a number of different genes can be analyzed in one cell preparation. With the total RNA from ∼1,000 cardiomyocytes or 150 capillary fragments (each containing 6–15 endothelial cells), >30 PCR runs can be performed. It is likely that with minor modifications this method can also be applied to cell-specific RT-PCR analysis in other organs.
We thank Ibolya Bakos, Brigitte Burk, and Andrea Schubert for excellent technical help and Dr. Christian Peiser for the GAPDH-specific primers.
Address for reprint requests and other correspondence: R. Preisig-Müller, Institut für Normale und Pathologische, Physiologie der Universität Marburg, Deutschhausstrasse 2, D-35037 Marburg, Germany (E-mail:).
This work was supported by the Deutsche Forschungsgemeinschaft (Da 177/7-2) and by a grant from the Karl and Lore Klein-Stiftung.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
- Copyright © 1999 the American Physiological Society