Have bovine serum albumin during serum starvation to help the survive of the cells

I heard this trick in the lab meeting. "0.5% for Rat2 cells and 0.35% for MTLn3 cells. BSA is fatty acid free. Someone in the Burridge lab is also using this trick." Bovine serum albumin addition to the serum free may help to maintain the osmotic pressure or keep the membrane surface receptors busy? I have not used it for my experiments. I think it is a good thing to know.
Then, I tried to find the source of it by tracing Dr. Gregg Gundersen's paper. I still do not know where this trick started, but I did found some interesting experimental trick in one of his protocol.

Methods Enzymol. 2006;406:579-92.
Real-time centrosome reorientation during fibroblast migration.
Gomes ER, Gundersen GG.

Department of Anatomy & Cell Biology, Columbia University, New York, NY, USA.

The centrosome is positioned between the nucleus and the leading edge of many types of migrating cells. Cdc42 regulates this centrosome reorientation through its effectors Par6 and MRCK. Using time-lapse microscopy of live cells, the mechanisms and kinetics of centrosome reorientation can be studied. In this chapter, we describe a modification in the standard wound healing assay that allows the study of signaling pathways involved in centrosome reorientation and other polarization events that occur before cell migration. We also describe a method for visualization of centrosome reorientation by time-lapse microscopy using NIH 3T3 fibroblasts stably transfected with GFP-tubulin.

PMID: 16472689

Some experimental tips for live-cell imaging are collected here.

• Durg selection transient transfected cells for stable expression clone. Transfect cells with GFP-tubulin vector using lipofectamine according to the protocol provided by the manufacture. Two days after transfection, plate cells at low density with DMEM medium supplemented with 1 mg/ml G418. Once colonies are visible by direct observation of the plates (5–7 days after transfection), select individual colonies using cloning cylinders and transfer them to larger plates for expansion. Reduce G418 concentration to 0.5 mg/ml at this stage. When clones have been expanded, check for GFP-greek small letter alpha-tubulin expression by plating cells on acid-washed glass coverslips ... The level of GFP-tubulin relative to endogenous alpha-tubulin is <5% in the selected clone (3T3-GFPTub) (Fig. 2). Multiple clones should be checked for use in cell assays to ensure that the behavior being studied is typical.
• Stimulate cell polarization and migration in wound assay. The mitogenic lipid lysophosphatidic acid (LPA) is a major serum factor required for MT polarizations in response to wounding (Cook 1998 and Palazzo 2001). LPA induces the same MT reorganizations as serum, yet LPA alone is not sufficient to stimulate cell migration (Gomes et al., 2005). Thus, addition of LPA only activates cell polarization and allows the study of polarization events independently from those of active cell migration. To induce both cell polarization and migration responses, serum can be added to the starved cells (Gomes 2005, Gundersen 1988 and Gundersen 1994).
• Why serum starvation. Because serum factors are necessary for triggering the responses to wounding, serum starvation also allows a separation of the physical act of wounding from the triggering of polarization and migration responses. This has a number of advantages for studying cell polarization and migration, particularly in studying the very early polarization and migration responses. We have also found that serum-starved cells are flatter and provide better subjects for optical imaging. However, starved cells are sensitive to changes in medium; therefore, if the starvation medium is replaced by recording medium without LPA or serum, cells can detaching rapidly (15–30 min).
• Preparation of LPA Stock Solution. Add 2.3 ml of a sterile 10 mM HEPES, pH 7.4, 10 mg/ml BSA (fatty acid free, Sigma), 100 mM NaCl solution to 1 mg of LPA (Sigma). Mix well. Aliquot and freeze at −80°. Frozen stocks are good for at least 2 months. Thawed aliquots remain usable for 24 h.
• Preparation of Starved Cell Monolayers and Wounding. Allow cells to grow until just confluent (48 h). Fill three 6-cm dishes with 8 ml of DMEM without serum. Fill 35-mm dishes with 1 ml of DMEM without serum (or each well of a 6-well plate). Using a sterile forceps, transfer each coverslip (containing the confluent monolayer of cells) serially to the 6-cm dishes containing DMEM. Place the coverslip in the 35-mm dish and starve the cells for 48 h. Using a sterile jeweler's screwdriver (tip width 1–2 mm), wound the monolayer with an even, continuous movement. Multiple wounds can be made on a single coverslip. Replace cells in incubator and let recover for 20 min. Treat cells with 2 μM LPA to induce centrosome reorientation and MT stabilization or add serum (1–5%) to stimulate migration. Centrosome reorientation, MT stabilization, and cell migration can be assessed with fixed cell assays. Wound edge cells can be microinjected after recovery from the wound. Proteins can be microinjected into the cytoplasm, and DNA can be injected into the nucleus. Typically, we find that DNA expression vectors based on CMV promoters allows for efficient protein expression in 1–2 h after microinjection. Because cells will still polarize and migrate if LPA or serum are added up to 8 h after wounding, this allows proteins to be expressed before triggering wound responses.
• Recording media. 3T3-GFPTub cells are plated in specific imaging chambers, and modified medium (“recording medium”) is used for starvation and for maintaining the cells while imaging. [HBSS (Gibco) containing essential and nonessential MEM amino acids (Gibco), 2.5 g/l d-glucose, 2 mM glutamine, 1 mM sodium pyruvate, and 10 mM HEPES, pH 7.4.] DMEM medium is not suitable for fluorescence microscopy because it contains riboflavin that interferes with GFP fluorescence microscopy. Riboflavin has an excitation peak at not, vert, similar 450 nm and an emission peak at not, vert, similar 530 nm, which is similar to enhanced GFP excitation and emission peaks (480 nm and 530 nm, respectively). We also leave out the phenol red that is commonly included in media, because it can absorb light, and this is important for fluorochromes such as rhodamine or DsRed. The recording medium is buffered with HEPES, not bicarbonate, because most microscope chambers are not equipped with a CO2-enriched atmosphere. The imaging chamber must be covered, because evaporation is significant at 37°, especially during long recordings.
• Live-cell imaging. Starve cells in 4 ml recording medium for 24 h. Wound the monolayer and allow to recover for 20 min in the incubator. Transfer the imaging chamber to the microscope stage. To induce centrosome reorientation, add 100 μl of 80 μM LPA drop-wise (2 μM final concentration) or 40 μl of serum (1% final concentration).
• Photodamage. Long exposure times can result in photodamage of the cells without detectable photobleaching of the GFP-tubulin. Photodamage is recognized when cells bleb, MTs depolymerize or, more subtly, do not respond to LPA or serum addition. To minimize photodamage, reduce light exposure as much as possible while searching for cells and during imaging. It is helpful to dark-adapt one's eyes by waiting not, vert, similar5 min in complete darkness before searching for cells under the microscope. Also, we have found that the use of neutral density filters in combination with longer exposure times generates less photodamage than short exposure times without neutral density filters.

Different cells need different condition. This is a very good starting protocol for live-cell imaging.