Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jul 14.
Published in final edited form as: Methods Mol Biol. 2025;2956:23–33. doi: 10.1007/978-1-0716-4706-6_3

Isolation and Enumeration of Circulating Endothelial Cells with Antibody-Labeled Staggered Herringbone Microfluidic System

Mahedi Hasan 1, Angel Ricardo Lerma 1, Taslim A Al-Hilal 1
PMCID: PMC13360855  NIHMSID: NIHMS2189117  PMID: 40859072

Abstract

Circulating endothelial cells (CECs) are a key indicator of cardiovascular dysregulation. It is possible to efficiently isolate live CECs from patient blood samples for subsequent analysis by using microfluidic systems. Antibody-based methods ensure efficient and high-accuracy isolation of CECs. By performing this isolation under septic conditions, it is possible to collect viable CECs, which can then be cultured for further study. In this work, we detail the process of isolating and enumerating CECs from patient blood using biotinylated x-hu anti-CD31 antibody-coated staggered herringbone (μ-HB) devices.

Keywords: Circulating endothelial cell, Cardiovascular dysfunction, Microfluidic cell isolation, Herringbone device, Circulating biomarkers, Antibody-coated microfluidics

1. Introduction

Circulating endothelial cells (CECs) have been recognized as essential biomarkers in cardiovascular diseases [1]. Research over the last decades has explored CECs’ role in endothelial dysfunction as agents that provide valuable information to the medical field for a better understanding of prognosis, diagnostics, and therapeutic improvements [2]. It has been stated that CECs are rare in the bloodstream in healthy individuals. However, patients with abnormal conditions, such as metabolic diseases, inflammation, cardiovascular diseases, and infections, present higher CEC levels, which might indicate severe damage to the endothelium [3–5]. In whole blood, CECs are considered indicators of vascular activation and damage. CECs also serve as noninvasive markers of endothelium health [6]. Studies have independently associated the presence of CECs in blood with a threefold increase in adverse cardiovascular events, even after adjusting for factors such as age, race, coronary artery disease, and diabetes. The final model confirmed CEC >20 as a strong predictor of cardiac dysregulation [5].

Microfluidic chips generally allow faster analyses, multiplexed assays, and high throughput due to shorter separation and reaction times [7]. In cell isolation, microfluidic systems allow for precise control of fluid flow and cell manipulation, which results in a higher yield and viability of targeted cells with minimal processing time [8, 9]. Moreover, these systems require smaller sample volumes, making them particularly advantageous when working with limited patient samples [10]. Specifically, the microchannel system of a μ-HB chip can be adapted to isolate CECs. μ-HB chip design creates chaotic flow patterns that enhance fluid mixing, promoting efficient cell surface interactions and improving the capture rate of uncommon cells like CECs [11, 12]. The integrated design of microfluidic devices enables real-time monitoring and analysis of isolated CECs, facilitating immediate downstream applications and enhancing the overall efficiency and effectiveness of the isolation process [13, 14].

The challenge of effective cell isolation has impeded the utilization of CECs in cardiovascular disease research. In this study, we employ CD31 antibody-coated μ-HB chips to isolate CECs from patient blood samples. The μ-HB device used in this work measured 55 mm in length, 8 mm in width, and 0.1 mm in height, featuring 320 herringbone patterns. Each pattern is 50 μm wide and 50 μm tall, with a 40 μm spacing between adjacent patterns. The μ-HB pattern optimizes cell capture by enhancing contact between the cells and the capture antibodies. This method can facilitate earlier and more precise detection of CECs as markers of cardiovascular disease.

2. Materials

Unless otherwise mentioned, always use deionized (DI) water to make any aqueous solution. When working with blood samples, follow all guidelines from the appropriate authority.

2.1. Mixing PDMS Solution

Mix Sylgard 184 base (Part A) with Sylgard 184 curing agent (Part B) in a 10:1 ratio (see Note 1).

2.2. Preparing 2 mg/mL PDDA

In this work, the PDDA was used as a solution with a concentration of 20 wt. % in H2O. Mix 0.01 mL or 10 μL from the PDDA solution with 4.95 mL of DI water to make a 2 mg/mL PDDA solution (see Notes 2 and 3).

2.3. Preparing 2 mg/mL Biotin–Alg

Weigh the lyophilized biotin–Alg. For every 2 mg of biotin–Alg, add 1 mL of DI water, and prepare the solution using a vortex mixer (see Note 4).

2.4. Preparing 50 μg/mL Streptavidin

When working with solid lyophilized streptavidin, it needs to be reconstituted before use. If the vial contains 5 mg of streptavidin, add 100 mL of DMSO/PBS to make a solution of 50 μg/mL (see Note 5).

2.5. Preparing 20 μg/mL Fluorescent-Conjugated Primary Antibody Solution

If the vial has 1 mg of antibody, add 50 mL of DMSO/PBS to make a solution of 20 μg/mL (see Note 6).

2.6. Preparing 0.5 IU Heparin Solution

From a 0.5 mg/mL heparin solution, take 20 μL for each mL to make a 0.5 IU heparin solution (see Note 7).

2.7. Preparing Whole Blood Samples for CEC Isolation

Mix the whole blood in a 1:1 ratio with the saline. In that solution, add the necessary amount of heparin to prevent aggregation (see Note 8).

2.8. Preparing 0.1% Bovine Serum Albumin (BSA) in PBS

Add 20μL of 5% BSA to 9.98 mL of PBS to make 0.1% BSA in PBS.

2.9. Preparing Staining Mixture

To prepare the antibody cocktail mixture for staining, add 5 μL of anti-CD146 from 20 μg/mL, 5 μL of anti-CD45 from 20 μg/mL, and 5 μL of DAPI with 5 μL of PBS. The final concentration of each antibody is 5 μg/mL.

3. Methods

The fabrication and coating process for all PDMS μ-HB chips takes approximately 1 day. Therefore, it is essential to plan accordingly to ensure the devices are prepared and readily available when working with patient samples. Once the devices are coated with appropriate capture antibodies, they can be stored at 4 °C for up to 1 week in aseptic conditions before they are used for isolating cells. Use at least three μ-HB devices for isolating CECs from each patient sample. If the intended target is to isolate and culture CECs, all operations should be performed under aseptic conditions once the μ-HB device fabrication is completed. After fabricating the μ-HB device, it should be sterilized under UV instead of flowing ethanol or other toxic chemicals, as they tend to leave residues inside the chip that can harm the coating process or cell capture.

3.1. All PDMS Microfluidic μ-HB Device Fabrication

3.1.1. Making PDMS μ-HB Structure

  1. Weigh the Sylgard 184 base (Part A) and the Sylgard 184 elastomer (Part B) in a 10:1 ratio using a weighing instrument, and take it in a plastic/glass jar (see Note 9) (Fig. 1a).

  2. Mix the weighted base and elastomer solutions with a stirrer (see Note 10).

  3. Put the mixture in a vacuum desiccator, and degas until all air bubbles are removed from the PDMS mixture (see Note 11) (Fig. 1b).

  4. Pour the mixture into the master mold (see Note 12). Bake it for at least 90 min (about 3 h) at 60 °C (see Note 13) (Figs. 1c, d).

  5. Extract the PDMS from the HB master mold, and cut excess areas to get the final device (see Note 14) (Fig. 2a).

  6. Apply transparent scotch tape to the side of the μ-HB structure to prevent dust particles from accumulating on the surface.

Fig. 1.

Fig. 1

Open in a new tab

Fabricating μ-HB structure. (a) Mix the Sylgard Part A (base) and Part B (curing agent). (b) Degas the mixture till it is completely bubble-free. (c) Pour the degassed mixture into the μ-HB master mold. (d) Put the μ-HB master mold with PDMS mixture in the oven to bake at 60 °C for at least 2 h

Fig. 2.

Fig. 2

Open in a new tab

Making μ-HB device from μ-HB structure and PDMS base. (a) Extracting the μ-HB structure after curing. (b) μ-HB structure and PDMS base alignment. (c) Plasma treatment and fabricating the μ-HB device. (d) Step-by-step coating of the μ-HB device

3.1.2. Making PDMS Base

  1. The degassed PDMS mixture can make the PDMS base. When using a 150 mm2 petri dish, adding 45 g of mixture can give a PDMS base of 2–3 mm thickness.

  2. Once the mixture is uniformly distributed in the dish, degas it again to remove any trapped air while pouring the viscous PDMS solution.

  3. Bake it for 90 min (see Note 15).

  4. Cut 75 × 25 mm area PDMS from the cured samples for the base (see Note 16).

  5. Apply transparent scotch tape on both sides of the PDMS base to prevent dust particles from accumulating on the surface.

3.1.3. Making All PDMS μ-HB Devices

  1. Layer the μ-HB structure and the PDMS base, ensuring that the PDMS base completely covers the μ-HB structure (Fig. 2b).

  2. Mark the areas for the inlet and outlet holes to be punched on the PDMS base (see Note 17) (Fig. 2b).

  3. Punch the holes on the PDMS base with a 1.5 mm biopsy punch.

  4. Peel off the tapes from the μ-HB structure and PDMS base.

  5. Put them in the plasma chamber for treatment on a quartz glass slide (see Note 18). Put the structure side facing up.

  6. Plasma treat for 90 s and bring them out of the instrument.

  7. Once out, align the μ-HB structure with the holes on the PDMS base, and attach them quickly to fabricate the μ-HB devices (see Note 19) (Fig. 2c).

  8. Store the prepared μ-HB devices in a dry place (see Note 20).

3.2. Isolation of CECs with μ-HB Device

3.2.1. Coating the μ-HB Devices with Capture Antibody

  1. Coat the surface of the μ-HB devices with PDDA of 2 mg/mL concentration. Leave it inside the chip for 5 min to allow the adsorption of PDDA by PDMS (see Note 21).

  2. Aspirate out the excess PDDA using a syringe. Then, wash the chip with excess DI water (> 300 μL) to remove any residue (see Note 22).

  3. Add 2 mg/mL of the biotin–Alg solution after washing. Leave it inside the chip for 5 min for PDMS to adsorb biotin–Alg.

  4. Aspirate out the excess biotin–Alg using a syringe. Then, wash the chip with excess DI water to remove any residue.

  5. Repeat steps 1–4 four more times (a total of five times) to get the layer-by-layer assembly (see Note 23).

  6. After washing the chip, coat it with 50 μg/mL streptavidin solution, and store it at 4 °C for overnight coating.

  7. After overnight coating, thoroughly aspirate the streptavidin and wash with DI.

  8. Coat the chips with biotinylated x-h CD31 antibody for at least 4 h before starting the capture process (see Note 24) (Fig. 2d).

  9. Once the capture antibody coating is done, these μ-HB devices can be used for CEC isolation or stored at 4 °C for later use.

3.2.2. Isolating CECs with Antibody-Coated μ-HB Devices

  1. Dilute the blood sample using saline in a 1:1 ratio in 5 mL epi tubes (see Note 25).

  2. Add an appropriate volume of low-molecular-weight heparin to the blood sample so that it may have 0.5 IU activity.

  3. Punch 1.0-mm-diameter holes through the cap of the epi tube (see Note 26).

  4. Soak the 1.5 mL tubes in 0.5 IU heparin solution before flowing the blood through them.

  5. Insert the 1.5 mL tubes through the epi tube caps.

  6. Place the blood-containing epi tube on a shaker so they do not coagulate/aggregate (see Note 27).

  7. Connect 1.5 mL tubes with 3 mL syringes and place them in the syringe pump.

  8. Connect the 1.5 mL tubes from the epi tubes to the inlet and the tubes from the syringes to the outlet channels.

  9. Flow at least 200 μL of blood through each of the μ-HB devices at a flow rate of 4 μL/min using the withdrawal function of the syringe pump (see Notes 28 and 29) (Fig. 3a).

  10. Once CEC capturing is complete, discard all remaining blood, 1.5 mL tubes, and 5 mL epi tubes following the proper guidelines.

  11. Wash the μ-HB devices as quickly as possible once the CEC capture is completed at a 7 μL/min flow rate with at least 200 μL 0.1% BSA containing PBS (see Note 30).

Fig. 3.

Fig. 3

Open in a new tab

Isolation and enumeration of CECs from whole blood sample. (a) Scheme showing CEC isolation setup. (b) On-chip imaging to distinguish CECs from immune cells. (c) FACS gating to distinguish CECs from immune cells

3.3. Staining and Counting CECs

3.3.1. On-Chip Staining and Counting of CECs with Antibody-Coated μ-HB Devices

  1. After washing, add 4% PFA to fix the cells on the μ-HB device at a flow rate of 1 μL/min for 40 min (see Note 31).

  2. Wash the excess PFA with 200 μL of 0.1% BSA containing PBS (see Note 32).

  3. Permeabilize the cells using 0.3% Triton X-100 and block with 5% BSA for 20 minutes.

  4. Wash any excess solution with 0.1% BSA containing PBS.

  5. Add 20 μL of the prepared staining mixture of anti-CD146 (5 μg/mL), anti-CD45 (5 μg/mL), and DAPI to each of the μ-HB devices.

  6. After overnight incubation, these chips can be imaged to determine the number of CECs captured.

  7. Cells with DAPI+/CD146+/CD45− are CECs, while cells with DAPI+/CD146−/CD45+ are immune cells (Fig. 3b).

3.3.2. Staining of CECs for FACS

  1. To perform FACS analysis after initial washing, extract the cells using a high flow rate of 15 μL/min to release and collect them (see Note 33).

  2. Fix the cells with PFA for 20 min.

  3. Perform centrifuge at 2.2 k rpm for 5 min.

  4. Aspirate the supernatant, leaving the precipitated cells in the tube with a slight amount of media.

  5. Add the mixture of 0.1% Triton-X permeabilization buffer and 5% BSA, and leave it at 4 °C for 30 min.

  6. Wash following step 3.

  7. Add the antibody cocktail to the tubes, and incubate for at least 45 min in the dark at RT or leave it for overnight staining at 4 °C.

  8. Wash following step 3.

  9. Add FACS buffer and perform FACS analysis (see Notes 34, 35, and 36) (Fig. 3c).

4. Notes

1.

Using a disposable plastic jar is better since removing all traces of PDMS from a jar is almost impossible. Using a new plastic jar is recommended every time.

2.

The density of PDDA is 1.04 g/mL. The concentration of PDDA for the solution is 20 wt. % in H2O. So, 1 mL of the solution contains 0.208 g of PDDA. From there, it is possible to determine the volume of PDDA solution needed to make the final 2 mg/mL solution using the V1S1 = V2S2 formula. For making a 1 mL stock solution V1 = 1000 μL, S1 = 2 mg/mL, S2 = 208 mg/mL, V2 = ?. So, V2 = (V1 × S1)/S2. This formula can be used to make any volume of stock solution.

3.

The PDDA solution is viscous. Therefore, cut the pipette tip to ensure it can take the PDDA solution. Also, it is advisable to make at least 5 mL solution as that needs 50 μL of PDDA. It is easier to pipette out 50 μL of PDDA using a 200 μL tip compared to pipetting out 10 μL of PDDA using a 20 μL tip.

4.

While mixing, make sure there is no residue. It usually needs vortex mixing for a few minutes to ensure the biotin–Alg completely dissolves in DI water.

5.

Always check with the manufacturer to determine which chemicals should be used for reconstituting their lyophilized products. For this purpose, unconjugated streptavidin should be used. Otherwise, if a fluorescence agent of a similar excitation– emission wavelength is used in CEC isolation, it will interfere with imaging.

6.

Use x-Hu antibodies if this system is used to identify cells from human samples.

7.

Make the heparin solution with 0.5 mg/mL concentration. That way, every 1 mL would have 50 IU of activity. From there, take 20 μL for every 1 mL of blood to make a heparin solution of 0.5 IU. A heparin solution with higher IU can lead to blood cell shape deformation.

8.

If necessary, add even more saline to the blood while ensuring it is not affecting the cells in the blood. Also, it is essential to mark down the dilution ratio as that will be necessary in the final calculation for CEC isolation.

9.

The weight does not need to be exactly 10:1. Also, the 10:1 ratio can be w/w or v/v. So, you can mix a 9 g base with 1 g elastomer or a 9 mL base with 1 mL elastomer. However, it should not exceed ±0.2 g. If too much elastomer is added, the PDMS will be more rigid and brittle. If too little elastomer is added, the PDMS will be very flexible. This is due to the amount of cross-linking.

10.

Stir using a glass rod or small plastic spatula until you do not feel any resistance while stirring. Once the base and elastomer are mixed thoroughly, they should look foamy and have many bubbles.

11.

It takes approximately 20 min to degas a PDMS mixture with 20 g weight. Depending on the weight and trapped air quantity, it might take longer to degas the PDMS mixture completely.

12.

For an all PDMS μ-HB device, it is essential to make the HB structure as thin as possible. Adding ~15 g of PDMS mixture for each HB structure can ensure that the thickness of the final structure is around 0.3–0.5 mm.

13.

Baking can also be done for extended periods. Once the PDMS solidifies, it will not undergo significant shrinkage even when baked for much longer.

14.

Be careful not to break the PDMS structure when pulling it away from the silica master mold. Also, do not punch any holes in the μ-HB structure.

15.

PDMS device and base preparation can be done simultaneously to save considerable time in device preparation.

16.

The base does not need to be precisely 75 × 25 mm. However, it must be bigger than the μ-HB structure (55 × 8 mm) to seal and create the μ-HB device effectively.

17.

Mark the inlet and outlet with easily distinguishable symbols. For instance, one can mark the inlet with a circle and the outlet with a plus symbol. Also, put the same symbols on the μ-HB structure. That way, aligning the sides after plasma treatment to make the final device is easy.

18.

Plasma treats an identical number of μ-HB structures and PDMS bases simultaneously.

19.

If a structure is left without attachment after plasma treatment, the surface might be deactivated, and the device bonding will not be remarkably effective.

20.

Once the μ-HB devices are made, they can be stored for an extended time and can be used up to 3 months later.

21.

Use a 1 mL syringe with the 1.5 mL tube for coating with PDDA.

22.

Use 3 mL syringes with the 1.5 mL tube for aspiration and washing with DI.

23.

If any bubbles form during the coating steps, tap on the μ-HB devices until all bubbles are cleared away so that uniform surface coating can be obtained.

24.

If the coating is done at RT, 4 h should be enough to ensure biotin–avidin binding if the coating is done at 4 °C.

25.

Mix at least 500 μL of blood with 500 μL of saline for each μ-HB device.

26.

The 1.00 mm radius hole ensures that the tubes are sealed. This is important for withdrawing blood from the tubes without allowing air to infiltrate. The tubes need to be in a place so that even if there is only a tiny amount of blood in the tube, it can still flow through. Putting the tubes too much further in would result in losing adequate blood that can be tested.

27.

Once heparin is added, blood does not coagulate immediately. It is possible to run the blood samples for at least 1 h and 20 min before aggregates appear under the microscope flowing inside the μ-HB device.

28.

It is recommended that a flow rate no higher than 5 μL is used, as when the flow rate is increased, more CECs tend to escape capture. At the same time, a prolonged flow rate tends to take too long, and the blood cells start to aggregate, making it challenging to isolate CECs.

29.

If the μ-HB devices are placed under the microscope, it is possible to image CEC isolation in real time.

30.

Blood will coagulate if allowed to stay in the μ-channels of the HB structure. So, all remaining blood must be washed away immediately. However, do not use an excessive flow rate; otherwise, captured cells will wash away.

31.

It is best to add PFA slowly since it is viscous and might detach captured cells. Each μ-channel can hold a maximum of 20 μL solution. So, at a flow rate of 1 μL/min, it would take at least 20 min to cover the μ-HB device with PFA.

32.

Steps after fixing with PFA can be carried out using a pipette instead of a syringe pump.

33.

Collect the cells in a 1.5 mL epi tube. It is easier to work with such a small volume in small tubes.

34.

Add enough FACS buffer according to the instrument requirements. Otherwise, the instrument will not be able to run the sample analysis. The BD FACS Accuri C6 Plus instrument requires at least 200 μL buffer to count the cells.

35.

Run blank solution to compare with the sample and distinguish signals from the CECs. Gates for FACS analysis should be taken based on the comparison with the control.

36.

If 0.5 mL of blood was taken, the initial dilution with saline for blood was 1:1, including a two-time multiplication to the final CEC count.

References

  • 1.Brett VE, Dignat George F et al. (2024) Circulating endothelial cells in pathophysiology. Curr Opin Hematol 31(3):148–154. 10.1097/MOH.0000000000000814 [DOI] [PubMed] [Google Scholar]
  • 2.Boos CJ, Lip GY, Blann AD (2006) Circulating endothelial cells in cardiovascular disease. J Am Coll Cardiol 48(8):1538–1547. 10.1016/j.jacc.2006.02.078 [DOI] [PubMed] [Google Scholar]
  • 3.Kumboyono K, Chomsy IN, Firdaus DH et al. (2022) Protective cardiovascular benefits of exercise training as measured by circulating endothelial cells and high-density lipoprotein in adults. J Taibah Univ Med Sci 17(4): 701–706. 10.1016/j.jtumed.2021.12.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.van der Feen DE, Bartelds B, de Boer RA et al. (2019) Assessment of reversibility in pulmonary arterial hypertension and congenital heart disease. Heart 105(4):276–282. 10.1136/heartjnl-2018-314025 [DOI] [PubMed] [Google Scholar]
  • 5.Mohandas R, Diao Y, Chamarthi G et al. (2021) Circulating endothelial cells as predictor of long-term mortality and adverse cardiovascular outcomes in hemodialysis patients. Semin Dial 34(2):163–169. 10.1111/sdi.12943 [DOI] [PubMed] [Google Scholar]
  • 6.Soltero EG, Solovey AN, Hebbel RP et al. (2021) Relationship of circulating endothelial cells with obesity and Cardiometabolic risk factors in children and adolescents. J Am Heart Assoc 10(1):e018092. 10.1161/JAHA.120.018092 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Shields CW 4th, Reyes CD, López GP (2015) Microfluidic cell sorting: a review of the advances in the separation of cells from debulking to rare cell isolation. Lab Chip 15(5): 1230–1249. 10.1039/c4lc01246a [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Wang C, Qiu J, Liu M et al. (2024) Microfluidic biochips for single-cell isolation and single-cell analysis of multiomics and exosomes. Adv Sci (Weinh) 11(28):e2401263. 10.1002/advs.202401263 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Tian Y, Hu R, Du G et al. (2023) Microfluidic chips: emerging technologies for adoptive cell immunotherapy. Micromachines (Basel) 14(4): 877. 10.3390/mi14040877 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Enders A, Grünberger A, Bahnemann J (2024) Towards small scale: overview and applications of microfluidics in biotechnology. Mol Bio-technol 66(3):365–377. 10.1007/s12033-022-00626-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Stott SL, Hsu CH, Tsukrov DI et al. (2010) Isolation of circulating tumor cells using a microvortex-generating herringbone-chip. Proc Natl Acad Sci USA 107(43): 18392–18397. 10.1073/pnas.1012539107 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Binsley JL, Myers TO, Pagliara S et al. (2023) Herringbone micromixers for particle filtration. Biomicrofluidics 17(1):014106. 10.1063/5.0134431 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Hama B, Mahajan G, Fodor PS et al. (2018) Evolution of mixing in a microfluidic reverse-staggered herringbone micromixer. Microfluid Nanofluid 22:54. 10.1007/s10404-018-2074-0 [DOI] [Google Scholar]
  • 14.De Jonghe J, Kaminski TS, Morse DB et al. (2023) spinDrop: a droplet microfluidic platform to maximise single-cell sequencing information content. Nat Commun 14(1):4788. 10.1038/s41467-023-40322-w [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES