In Vitro Matrigel Assay 205

205

From: Methods in Molecular Medicine, Vol. 46: Angiogenesis ProtocolsEdited by: J. C. Murray © Humana Press Inc., Totowa, NJ

14

In Vitro Matrigel Angiogenesis Assays

M. Lourdes Ponce

1. IntroductionA variety of in vivo and in vitro methods have been used to study angiogen-

esis, the process of blood vessel formation. Two widely accepted but techni-cally difficult assays include the cornea implant assay and the chickchorioallantoic membrane assay. The cornea assay requires special equipmentand a skilled person to implant beads containing the test compound in the eyesof animals; only a small number of samples can be tested due to cost and time.The chorioallantoic membrane assay requires a large number of samples onaccount of the variability of the system and its difficulty in quantitation. In ourlaboratory, we have developed a quick and highly reliable method for testingnumerous compounds for angiogenic and/or antiangiogenic activity. Themethod is based on the differentiation of ECs on a basement membrane matrix,Matrigel, derived from the Engelbreth-Holm-Swarm tumor (1). ECs fromhuman umbilical cords as well as from other sources differentiate and formcapillary-like structures on Matrigel in the presence of 10% bovine calf serum(BCS) and 0.1 mg/mL of endothelial cell growth supplement (ECGS) (2),which is a mixture of both acidic and basic fibroblast growth factor (Fig. 1,Panel C). The formation of tube-like vessels under these conditions can beused to assess compounds that either inhibit or stimulate angiogenesis. Inthe assay, substances that affect angiogenesis, such as the laminin-1 peptidecontaining the IKVAV sequence, disturb the formation of capillary-like struc-tures and the distinctive morphological cell characteristics resulting are indicativeof the potential activity of the compound (Fig. 1, Panel D). However, it can notbe determined from the morphological characteristics whether a compound isangiogenic or antiangiogenic.

206 Ponce

Tubes do not form when both serum and ECGS concentrations are reduced(Fig. 1, Panel A). This allows one to determine whether a substance is angio-genic. Under these circumstances human umbilical vein endothelial cells(HUVEC) do not completely differentiate; instead, they form incomplete, shorttube-like structures. In the presence of angiogenic compounds, such as fibro-blast growth factor or hepatocyte growth factor, the ECs differentiate into well-defined tube-like structures (Fig. 1 Panel B). An additional advantage of theseassays is that they can be scaled down and 48- or even 96-well plates can beused if many compounds need to be tested or their quantities are limited. Theuse of either one or both of these assays can help to quickly and efficientlyidentify compounds with angiogenic or antiangiogenic activity. Using both ofthese assays, a number of new compounds that affect angiogenesis have beenidentified including hepatocyte growth factor (HGF), haptoglobin, estrogen,IP-10, and numerous laminin-1 peptides among others (3–6). It should be notedthat the activity in all cases should be confirmed by an additional assay.

Fig. 1. Endothelial cells on Matrigel. HUVEC were plated on Matrigel in the pres-ence of HUVEC media containing either 5% (A,B) or 10% BCS (C,D). Cells incu-bated with 25 ng/mL of hepatocyte growth factor (HGF) formed tubes in serumreduced media (B) as compared to their control (A). Cells that were plated with0.1 mg/mL of a laminin-1 peptide that contains the IKVAV sequence did not formcomplete tube-like structures (D) as compared to their negative control devoid of pep-tide but containing 10% BCS (C). Original magnification ×5.

In Vitro Matrigel Assay 207

2. Materials1. Matrigel (Sigma Chemical Co., Saint Louis, MO or Becton Dickinson, Inc., Ru-

therford, NJ).2. 24-Well tissue-culture plates.3. Pipetmen and sterile tips.4. Human umbilical vein endothelial cells (HUVEC) prepared from umbilical veins

or commercially obtained from Clonetics Corporation, San Diego, CA (see Note 1).5. RPMI 1640 media (1 L, Life Technologies, Gaithersburg, MD).6. Table centrifuge.7. 50-mL Conical tubes.8. Hematocytometer.9. Trypsin-EDTA solution containing 0.05% trypsin, 0.53 mM ethylenediamino-

tetraacetic acid in Hanks’ Balanced Solution (Life Technologies, Gaithersburg, MD).10. Incubator with 95% air/5% CO2 mixture.11. HUVEC medium:

• 500 mL RPMI 1640• 100 mL of bovine calf serum (BCS), defined and supplemented (HyClone

Laboratories Inc., Logan, Utah)• 0.1 g ECGS (Collaborative Biomedical Products, Bedford, MA)• 2,500 U of heparin sodium (Fisher Scientific, Fair Lawn, NJ)• 5 mL of fungizone solution (Life Technologies, Gaithersburg, MD)• 1.25 mg gentamicin (Life Technologies, Gaithersburg, MD)• 5 mL of penicillin/streptomycin solution (10,000 U/mL of penicillin G sodium

and 10 mg/mL streptomycin sulfate in 0.85% saline; Life Technologies,Gaithersburg, MD).

12. Diff-Quik fixative and solution II (Baxter Scientific Products, McGaw, IL)

3. Methods3.1. Testing Angiogenic Compounds

1. Allow Matrigel to thaw on ice at 4°C; once thawed keep on ice at all times or itwill solidify at room temperature.

2. Under sterile conditions, coat 24-well plates with 300 µL per well of Matrigelwithout introducing air bubbles. Let the plates sit at room temperature for at least15 min to allow gelling of Matrigel.

3. Rinse a confluent 150-mm dish or equivalent containing 3–4 × 106 ECs in 5 mLof RPMI 1640 medium.

4. Add 5 mL of trypsin-EDTA solution and incubate until the cells detach from theplate (approx 5 min).

5. Place the cells in a 50 mL conical sterile tube and add 5 mL of HUVEC media toneutralize the trypsin.

6. Centrifuge the tube on a table top centrifuge for 5 min at 50g; remove the super-natant, tap the pellet gently and resuspend the pellet in 2 mL HUVEC medium.

7. Set aside a 0.2 mL cell aliquot as control.

208 Ponce

8. Count a 10 µL aliquot in a hemocytometer or use a Coulter Counter to determinethe total number of cells in the tube.

9. Adjust the volume with HUVEC media (containing 20% BCS) in order to have2 × 106 cells per mL (see Note 2).

10. To reduce simultaneously the concentration of serum to 5% and the number ofcells to 0.5 × 106 per mL, add an amount of RPMI 1640 media equal to threetimes the volume present in the tube.

11. Plate 192 µL of the cell mixture, containing 48,000 cells, into each Matrigel-coated well.

12. Add the test compound in a total volume of 8 µL (see Note 3).13. Controls that will contain 10% serum are similarly prepared from the aliquot that

was set aside, except that 1 × 106 cells per mL (instead of 2 × 106 cells per mL)are suspended in HUVEC media (20% serum) and are subsequently diluted todouble their volume with RPMI 1640 medium.

14. Gently swirl the dishes to evenly mix the test compound(s) with the cells.15. Incubate dishes overnight at 37°C in a 5% CO2/95% air incubator.16. The next day gently aspirate the medium from each well, add approx 200 µL of

Diff-Quik fixative per well and incubate for 30 sec.17. Aspirate the fixative and stain the cells for 2 min with Diff-Quik solution II that

has been diluted 1/1 with water (see Notes 4 and 5).18. Observe tube structures under a microscope and photograph.

3.2. Screening Compounds That Influence Angiogenesis

This method is a variation of the assay described in Subheading 3.1. andcan be used for screening compounds that either induce or inhibit angiogen-esis.

1. Follow protocol as in Subheading 3.1. up to step 8.2. Suspend cells in HUVEC media containing 20% BCS to yield 1 × 106 cells per mL.3. Dilute cell suspension with an equal volume of RPMI 1640 medium lacking

serum to reduce by half the amount of serum and the number of cells per mL.4. Continue assay as described in Subheading 3.1. Controls should include cells in

medium containing 5 and 10% serum (see Notes 6–8).

4. Notes1. HUVEC should not be older than 6 passages because they are primary cell cul-

tures and tend to lose their differentiated characteristics. They should always behandled with gloves since they are of human origin.

2. The amount of HUVEC media in which the cells are suspended in step 9 willvary from experiment to experiment according to the total number of cells har-vested from each plate.

3. Several concentrations of each compound should be tested at least in triplicate.Include 6 wells for controls (3 positive and 3 negative).

In Vitro Matrigel Assay 209

4. Staining of the Matrigel with Diff-Quik solution II for longer than 2 min or with-out previous dilution will excessively stain the Matrigel making it difficult toobserve the tubular structures.

5. Dishes containing stained cells can be stored wrapped with plastic for a week orlonger at 4°C. Pictures should be taken within a few days since the Matrigel tendsto dry up.

6. To perform assays on 48-well dishes, wells are coated with 200 µL of Matrigeland 24,000 cells are plated in 150 µL of medium. Ninety-six well plates are coatedwith 100 µL of Matrigel and plated with 14,000 cells in 100 µL of medium.

7. Unknown samples should be tested using both serum concentrations.8. HUVEC can form tubes even in the absence of serum in some commercially

available batches of Matrigel. It is recommended that a control experiment becarried out using low serum concentrations each time a new Matrigel lot is used.

References1. Kubota, Y., Kleinman, H. K., Martin, G. R., and Lawley, T. J. (1988) Role of

laminin and basement membrane in the morphological differentiation of humanendothelial cells into capillary-like structures. J. Cell Biol. 107, 1589–1598.

2. Grant, D. S., Kinsella, J. L., Fridman, R., Auerbach, R., Piasecki, B. A., Yamada,Y., et al. (1992) Interaction of endothelial cells with a laminin A chain peptide(SIKVAV) in vitro and induction of angiogenic behavior in vivo. J. Cell Physiol.153, 614–625.

3. Grant, D. S., Kleinman, H. K., Goldberg, I. D., Bhargava, M. M., Nickoloff, B. J.,Kinsella, J. L., et al. (1993) Scatter factor induces blood vessel formation in vivo.Proc. Natl. Acad. Sci. USA 90, 1937–1941.

4. Cid, M. C., Grant, D. S., Hoffman, G. S., Auerbach, R., Fauci, A. S., andKleinman, H. K. (1993) Identification of haptoglobin as an angiogenic factor insera from patients with systemic vasculitis. J. Clin. Invest. 91, 977–985.

5. Ponce, M. L., Nomizu, M., Delgado, M. C., Kuratomi, Y., Hoffman, M. P., Powell, S.,et al. (1999) Identification of endothelial cell binding sites on the laminin 1 chain.Circ. Res. 84, 688–694.

6. Malinda, K. M., Nomizu, M., Chung, M., Delgado, M., Kuratomi, Y., Yamada, Y.,et al. (1999) Identification of laminin 1 and 1 chain peptides active for endot-helial cell adhesion, tube formation, and aortic sprouting. FASEB J. 13, 53–62.