Membrane Biophysics - Science topic

Dear Peyman Fahimi ,

You are right lysolecithin is not a bilayer forming lipid so they will not form a proper membrane. The reference as pointed out by Shahin Sowlati Hashjin demonstrates that multiple effects occur see for example the following said by the authors “…experimental results were complicated by an apparent temperature dependent increase in the solubility of the micellar phase. This resulted in desorption from the surface, with a resultant increase in surface tension.”

Indeed I feel that you address quite a complicated issue, since I assume that different effects are intertwined (fluidity, viscosity, tendency to deformation of bilayers or in the case of surface tension monolayers, changes in adsorption and/or desorption etc.).

I think that the following papers is a good example of the multiple effects that can occur if you look at such complicated matters you are interested in:

Ghatee, M. H., Zare, M., Zolghadr, A. R., & Moosavi, F. (2010). Temperature dependence of viscosity and relation with the surface tension of ionic liquids. Fluid Phase Equilibria, 291(2), 188-194.

My educated guess would be that the temperature dependence is largely depending on acyl chain lengths while the extent of surface tension and elasticity is determined by the head groups.

Hope this answers your question somewhat.

Best regards.

Yes this clarifies a lot ;-)

Indeed this patent describes the use of an ester type compound (intrinsically hydrophobic) that is once loaded into liposomes is saponified into a hydrophilic one.

INTERESTED TOO

Method for preparation of uniformly oriented and sealed inside out vesicles (ISOv) from Escherichia coli cells was further developed and published by our group in Bogdanov, M., Heacock, P. and Dowhan, W.: A polytopic membrane protein displays a reversible topology dependent on membrane lipid composition. EMBO J. 21: 2107-2116, 2002. To prepare ISO vesicles Escherichia coli cells were ruptured using a French press at 560 kg/cm2 (8000 p.s.i. THIS PRESSURE IS CRUCIAL!). Orientation of membrane vesicles was verified by new vesicle sidedness assay by taking advantage of mapping of uniform topology of essential E. coli protein leader peptidase (Lep). Please see Materials and Methods and Supplemental material to EMBO J. 21: 2107-2116, 2002 manuscript for details of both protocols respectively.

I am interested in simulating the dynamics of membranes. However, the library you had sent me seems to focus on analyzing previously done simulations coming from MD programs.

@Alan I am not necessarily minimizing the energy. I am actually interested in using Monte Carlo simulations for studying dynamics and on the future some MD methods. What I really liked about Surface Evolver is that a surface can be defined and is messed and the energies are easily defined. None of the available MD programs I have seen have a triangulated surface that can be simulated and the energies provided by these programs care not meant for membrane simulations

Hi Bharadwaj,

The short answer to your question is that higher surface tension will make the membrane harder to bend.

The reason is: bending acts to reduce the effective area of the membrane while tension acts to increase the effective area. In simple cases, the balance of these two forces sets the equilibrium shape of the membrane.

There have been lots of studies related to membrane bending and membrane tension, one of my favorite is:

PMID: 12059401 DOI: 10.1103/PhysRevLett.88.238101

equation 1 is a good starting point to think about how different mechanical aspects of the membrane contributes to its free energy.

Best,

Zheng

Check: https://en.wikipedia.org/wiki/Nernst_equation

In principle, as simplification you can think it like this:

1) Lipid bilayer is impermeable to ions (though of course it really is not). 

2) Cell have ion pumps run by the energy of ATP cleavage, e.g. Na/K ATPase. Foe a moment, let us forget all the active pumps, and let us also forget that the ratio of transport for Na/K ATPase is roughly 3:2, and instead assume that it cleaves one ATP, and transports one Na+ ion out and one K+ ion in. This causes very different sodium and potassium concentrations in a cell. As such, it does not create any membrane potential, as one positive charge (in our example) is transported out for each positive charge that is transported in. Yet, energy is need to create the concentration difference. For sake on simplicity, let us say that we have 150 mmol/l Na+ outside and 3 mmol/l inside, and 3 mmol/l potassium outside and 150 mmol/l inside. 

3) Cell have selective ion channels (i.e. ion channels that let only particular ion go through) that can be open or closed.

3a) Let us now think that our cell that have lot of potassium in and a lot sodium out have potassium channels that are open. By random diffusion, it is more likely that the potassium ion will move from the high concentration volume (the inside of the cell) to the low concentration volume (the outside of the cell, although initially it is excepted to relative to concentration, so that initially 150 K+ are expected to go out for every 3 K+ that come in. So, this will lead into more positive charge going out than coming in. As a result, the inside of the cell will attain negative potential compared to the outside of the cell, or, equivalently, the outside of the cell will attain positive potential compared to the inside. It is just a standard we have chosen (wisely, since otherwise we would have problems if we had two cells) that the outside potential is zero, and hence the inside potential is the one that becomes negative. Since the potential is same in the volumes inside the cell and outside the cell, the potential gradient exists over the cell membrane, hence it is called membrane or, more exactly, transmembrane potential. When electric potential inside the cells becomes  sufficiently negative, then the net flow of potassium to the outside will cease, because the negative potential inside the cell or over the potassium ion channel will attract the positively charged potassium ions that are the ones that can flow through this open channel. This equilibrium can be calculated from the Nernst equation.

3b) Now that the actively created gradient of sodium and potassium and the flow of potassium out of the cells has created negative potential inside the cells compared to the outside, the other ion channels and transporters can use this negative potential. E.g. if an anion is present at equal concentrations, and an an ion channel opens, it will flow out of the cell until it reaches the Nernst equilibrium as well. Note that since the flow of the anion out of the cell will make the membrane potential less negative, more potassium will be allowed to flow out and this again tends to make the membrane potential more negative. The cells can also use transporters, e.g. if an anion that is present inside of the cells at greater concentrations than on the outside, then e.g. a transporter can couple to transport of this anion to the transport of let us say, two or three Na+ ions whose transport is favoured by both the concetration gradient and the membrane potential. So, the anion is transported into the cell, along with a few Na+ ion, and the membrane potential will tend to decrease, so that more K+ can again flow out of the cell so that a new combined equilibrium can be achieved. It is easy to see, nevertheless, that soon more and more potassium will flow out, and more and more sodium in, and so, even if the potassium channel conductance (i.e. their ability to allow for potassium flow in a time unit) is far larger than for any of the other ions, the system can only continue to run if cells actively use energy at Na+/K+ ATPase to maintain the potassium/sodium gradients inside and outside the cell.

Ok, thank you. I will try to explain from your point.

Hi Deepthi,

          You may want to have a look at this paper. It discusses the labeling protocols and constraints related to FM 1-43 labeling in endothelial cells.  

You might also want to consider using GFP tagged plasma membrane localizing probes(plasmid constructs)  such as CD36 , cadherin or another membrane marker for epithelial cells to study membrane dynamics

Hi there,

The case for the guanidinium of arginine is similar to the case of carboxylate in acidic aa (possibility of charge delocalization). They are both able to participate to salt bridge interaction which is actually made of both electrostatic interaction and hydrogen bonding contributions. Indeed global charge of arginine is delocalized between the 3 nitrogens of the guanidinium group but electrostatic interaction  may actually contribute to stabilize one of its resonance forms.

Hello Anna,

This is one of those topics which everyone under the sun has an opinion about.  There is no distinct method for calculating the EPT.  Since the ELISA was widely used to ascertain Ab presence, people have published ideas about calculating the titer.  I know you probably know this already, but the titer is simply expressing the point at which a dilution no longer has enough Ab to produce a response against antigen.  But there isn't a standard amount of antigen (too many antigens to have positive controls for each), and so you are basically trying to quantify the Ab response with respect to the available antigen.  Mainly to compare to other antibody-antigen responses, right?  The Frey paper from the 90's is probably as good as any.

I've always considered that when you picture Ab-Ag responses, as a reversible equilibrium reaction, then what do we say is a no response?  When the Ab response is not strong enough to have a signal generated in an assay?  Sounds kind of arbitrary to me, and unless the coating of the wells was the same, and the conjugation or secondary were the same, from one assay to the one three weeks from now, after bleed #2 or whatever, then the end-point is arbitrary at best.

I personally think that a four-point logistic regression of multiple dilutions (8 as a starting point) allows one to see the point at which the regression plot swings out of the linear range.  I think it is clear that is where the dilution is hampering the signal generation.  But again, the other parameters of the assay greatly affect where that point will be.

Good luck with your project.  Onnea!

In the following links you can find lipid coordinate and topology files for lipids:  http://wcm.ucalgary.ca/tieleman/downloads 

Using pdb2gmx command can be created mostly protein topology. Lipid (not all of them) and polymers topology is not included in the gromacs database. For your polymer case, just use one of the online automatic topology builders such as https://atb.uq.edu.au/index.py 

If you have further questions just let me know. 

Another way to concentrate the protein is to rebind it to a Ni²⁺ resin and elute it in a sharp band using a single step at high imidazole concentration.

Sodium fluoride, which does not absorb UV light, is a better choice if the protein is not especially sensitive to the choice of anion

Yep, give the waters KE that corresponds to that temperuture; the general recommendation is Berendsen Thermostats/Barostats rather than the accurate Nose-Hoover or Langevin, because of former having faster equilibriation time-scales.

Other than that I think you have done seems OK.

Best,  Prabhakar

Langmuir-Schaefer dips are rather straightforward to perform, although you need very precise equipment that might be very difficult to obtain (there's one at the PSCM, Grenoble, which I have used myself) and I guess this is the only technique that can really achieve asymmetric supported bilayers. However, your lipid composition (100% PS) can be tricky, and you may try to play around with pH and ionic strength for it to possibly work. Also, I'm not sure how long your membrane will really stay asymmetric (as stated above), as lipids can flip between the leaflets even without assisting enzymes. In theory one could try to use an electrochemical cell to stabilize the asymmetry (and there are people also working on such systems), but getting a completely asymmetric system with just PC on one side and PS on the other is probably impossible.

There are a lot of different spellings here. It is tortuosity. See Epstein, N. (1989), On tortuosity and the tortuosity factor in flow and diffusion through porous media, Chem. Eng. Sci., 44(3), 777– 779. [1]

Hi Marvin,

Nicolas and Norbert gave you the information you needed, The only thing I would advice is that if you have a nice large Ca current then it's better to use external Ca as a carrier as your experiment will be more physiologically relevant, Just remember to buffer internal Ca at low level with EGTA (e.g. use 0.5mM CaCl and 5mM EGTA).

Good luck

 If you want to check the effect of toxins on VOLTAGE-GATED Ca channels, then obviously you have first to activate those channels by depolarization. If you don't want to use electrophysiology (patch clamp, Ca imaging) you can activate channels by briefly applying high KCl (30-50mM). But first you need to load cells with some Ca indicator, e.g. Oregon Green BAPTA-AM and compare the changes in fluorescence before and after your toxin application. 

Please have a look at Computational Electrophysiology in GROMACS and also the following link..

Hope this helps

Thanks for clarifying and sorry I misunderstood your question.  I have not worked on membrane voltages of cells that are not neurons, but I would guess that the same principles of physical chemistry would apply to all cells in terms of establishing resting potentials.  One important issue is that some marine animals are osmoregulators, meaning that the concentration of ions in their blood differs from the concentration of ions in sea water.  Other marine animals are osmoconformers, meaning that the concentration of ions in their blood is nearly identical to that of the surrounding environment.  Thus, if the animals that you are interested in are osmoregulators, it might be that the actual concentration of ions in the blood differs from that of the outside environment.  In the end, the concentration of ions inside and outside of a cell, as well as the permeability of the cell membrane to each ion, will establish the membrane voltage.  If you are interested in the basic principles that underlie this, a neuroscience text discussing the Goldman equation and how resting potentials are established might be useful.  I hope this helps.  Good luck!

Thank you very much M. Cocera

Thanks Alexander S Krylov. I agree that there are certain methods available but I was more thinking about a direct and selective Raman spectroscopy method to identify the di-tyrosine bonds in the surface proteins.

two useful references:

1. Roehrl, M.H.A., J.Y. Wang, and G. Wagner, A general framework for

development and data analysis of competitive high-throughput screens for

small-molecule inhibitors of protein−protein interactions by fluorescence

polarization. Biochemistry, 2004. 43(51): pp. 16056-16066.

2. Huang, X., Fluorescence polarization competition assay: The range of

resolvable inhibitor potency is limited by the affinity of the fluorescent

ligand. J. Biomol. Screen., 2003. 8(1): pp. 34-38.

I also had the same doubt with fibronectin coating. You can put a small  glass coverslip or a round piece of parafilm on top of your droplet to make it flat an uniform.

Thank you so much for your elaborated response!  It is very helpful.

There are 2 reasons why you might need an ATP regenerating system: (1) depletion of ATP causes the reaction to slow down so that the initial rate of reaction is not maintained, or (2) accumulation of product ADP inhibits the reaction so that the initial rate is not maintained. If neither effect is observed, i.e. if the reaction progress curve is linear for the whole reaction time, then you do not need an ATP regenerating system.

A way to measure peptide / protein orientation in membranes is by attenuated total reflection IR spectroscopy (ATR-FTIR).

You use a IR transparent crystal as support for a lipid mono- or bilayer and mount it into a flow trough cell. You pump in your protein solution. You measure the light reaching the detector after multiple (total) reflections on its way through the crystal. The evanescent wave developing at each reflection is decaying within the first micrometer enough to see the membrane and proteins inserted or attached. You check the spectra for the amide bands of your helices and/or beta sheets.

With some assumptions on the system it is possible to calculate an order parameter for the helices and/or beta sheets of your protein in contact with the membrane.

Sorry , for prpteoliposomes I have no answer. Trial and error, the best

We have had success on a number of systems using styrene maleic acid co-polymers, which remove the need for detergents. Essentially they extract the membrane protein surrounded by its native lipids with the polymer acting like a nano disc scaffold.  Maybe they could help. See;

We did some studies (Ray et al, 1983) on the effects of stepwise removal of cholesterol on the structure and function of tightly sealed gastric microsomes. These  uniformly oriented microsomes are highly enriched  in H, K-ATPase that made the vesicles capable of accumulation H in exchange for K in presence of ATP. We observed that gradual removal of cholesterol by digitonin-treatment substantially inactivated both the ATPase activity and H-transport. Negative staining revealed large holes on the membranes resulting from cholesterol removal. Supplimentation of the depleted membranes with cholesterol restored the enzyme activity but not proton leakage, revealing the membrane structure was only partially restored. You get some idea on membrane ruffling by cholesterol from the studies.

Ray, TK, Nandi, J., Dannemann, A. and Gordon, GB (1983) Role of cholesterol in the structure and function of gastric microsomal vesicles. J. Celluilar Biochem. 21, 141-150

Hi John,

I would agree with Nicolas. The process of getting a good sample might take some optimisation, but it usually works. As long as you do not use the really large nano discs, you can feel relatively safe that you will only have one protein (monomer or whatever the native oligomeric state of your target protein is) per nano disc. There is simply not enough space in the smaller nano discs to fit much larger stuff in there. 

Hi Suchi, I haven't done this specific assay, but I have done other I125 binding studies.

If you don't get rid of free I125 from your prep, then you will get high background. You could try using NaI at ~100-500 fold molar excess over the ligand to se if that is the problem.

If you are using Chloramine-T or Iodogen to label your protein, then the problem could be oxidation of other amino acids that renders the protein inactive. Small proteins and peptides are more likely to be inactivated by iodine labeling.   

Is alpha-MSH Melanocyte-stimulating hormone? If so, then it has one Tyr that can be labeled, but it also has 1 Met, 1 His, 1 Trp and 3 Pro aa  that are susceptible to oxidation by Chloramine T and Iodogen.  

Hi Victoria,

we routinely measure the V-ATPase activity in membrane vesicles from plant tissue using a panel of inhibitors and conditions specific for V-ATPases. A bafilomycin A-sensitive ATPase activity is measured by detection of the released phosphate and the quench of acridine orange fluorescence due to H+ transport into the vesicles.

Addition of molybdate inhibits most of the phosphatases, vanadate blocks P-type ATPases and azide inhibits the F-type ATPases. If you add these inhibtors to your lysosomal fraction, use a pH of 7.5 and determine the released phosphate in the presence and the absence of bafilomycin A, the difference may well give you an exact estimate of the V-type ATPase activity.

You may find quite a number of recipies for the phosphate detection and the ATPase activity measurements by including plant sciences in your literature search. I just attached 2 as examples, an early original article and a more recent book chapter giving you the recipies for performing the assay.

Have fun with the measurements

Gerhard

I think you should use phosphatidyl choline (PC) containing liposomes for reconstituition.  We reconstituted the ion-transporting H, K-ATPase and Na, K-ATPase by PC  following inactivation by controlled delipidation of the plasma membrane under proper condition, where other phospholipids lipids were unsatisfactory. 

Many thanks for your input and suggestions. Really appreciate it.

We recently wrote a review about cortex and membrane tension that includes definitions for both as well as common measurement methods. Hope that helps!

I think your patchy staining is endocytosed membrane.  30mins at RT is quite a long time for the incubation.  I dilute rhoPE to around 1-3 ug/ml in PBS, put it on the cells for 10mins on ice, and get nicely uniform, very bright staining of the PM.  A couple of notes of caution though...  rhoPE is closer to orange than red, so will easily leak into your green channel.  Also, it has a tendency to photo-oxidize, which can have lots of unwanted effects on the membranes you're measuring.  So minimize the excitation light intensity as much as possible.

Both Jeffrey and James are correct. Generally liposomes are very difficult to look at via solution NMR due to the size and viscosity. If you have solid state NMR available it may be prudent to make use of the technique. In the following paper: Eur Biophys J (2008) 37:1031-1038 the changes in phospholipid membranes are examined after the introduction of a protein using MAS-NMR. While this may not be the answer you were looking for this may be a good paper to look at if you plan on doing any membrane protein studies using 2D crystals.

Yes. There are two methods, often used concurrently:

1) capacitance compensation circuitry

2) P/4 subtraction

Both serve to eliminate the capacitance charging transient, so that you can study the ionic currents alone.

The first approach is built into most modern commercial patch clamps. The capacitance charging transient is the current that flows in response to a voltage step applied from the patch pipette to the interior of the cell. The current enters the cell via the narrow aperture (so-called "series resistance") where the circular rim of the pipette connects to the cell interior, in whole-cell patch clamp configuration. Normally this current is applied by the measurement circuitry, so it is recorded along with any ionic (transmembrane) current. When the capacitance compensation circuitry is used and a voltage step command is applied, it is the compensation circuit that applies the current needed to charge up the membrane capacitor, relieving the measurement circuitry from having to supply that current. This results in the measurement circuitry reporting only the ionic current, without the capacitance charging current.

The second approach is based on the fact that the membrane capacitor roughly approximates a linear capacitor, which means that the capacitance transient that flows to charge the capacitor should have the same shape, but differ in size in proportion to the voltage step size. Simply put, to use this approach, one applies one or more small voltage steps -- too small to activate voltage-gated ion channels, and adds them up to get the predicted full sized transient -- and subtracts that full-sized transient from the current transient that results when a full-sized voltage step is applied. In the case of applying a full-sized voltage step, the current transient is the sum of both the ionic current and the linear capacitance transient, the latter of which can be subtracted out.

I will add some references later...

Reference?

Hi Sabya,

Thanks for your reply and nice explaination. But I think I'm looking for a more physical view of the lipid bilayer. I understand phosphatidylinositols are essential lipids mostly because they can interact with different peripheral proteins, and spontaneous curvature arised from protein-lipid interactions can be very complicated as you said.

But if we only look at the lipid bilayer, I think the shape of inidvidual lipids can also be interesting, at least biophysically. For example, the shape or spontaneous curvature of lipids will determine the prefered curvature of the monolayer. When two monolayers form a bilayer, the pressure profile should be somewhat related to the original shape of the monolayers. And the pressure profile across a lipid bilayer could also influence the insertion properties of proteins. The spontaneous curvature of different lipids in the wikipeida page such as for DOPE and DOPC, are mostly measured by x-ray diffraction from lipid hexagonal phases. So it should be a pure physical parameter of this lipid species, but again I agree with you about how protein-lipid interactions can make this much more complicated and also more meaningful.

Best,

Zheng

Thank you !

Did you measure the Tm from gel to liquid crystalline lamellar phase ???

Lipids, even uncharged zwitterionic lipids, have permanent dipoles. These can "feel" the AC field, they experiment a force and move the molecules along with it. This process is equivalent to vortexing and heating when making multilamellar vesicles: it makes it easier for water molecules to get into previously non-hydrated places, and so force the lipid molecules to come fully into contact with water. This in turn triggers the hydrophobic forces that will transform a previously dry and disordered lipid film into a bilayer, polar heads out, apolar chains inside. Regarding voltage and frequency values, these depend on the geometry of the system. You can check the literature for that. Right now I cannot remember precisely, but the numbers that come to mind are 10 Hz and 2 Volts for Pt wires separated by 4 mm. I can look it up if you need.

Mostly from x-ray from bent (hexagonal) phases

although their is a newer indirect method using fluorescence

Note that fluorescent lipids flip between much less frequently then natural lipids. Without looking into it in detail, I don't know how this affect their results.

Changes in the spontaneous membrane curvature can be approximated much easier by measuring the change in the phase transition temperature to the hexagonal (positive curvature) or inverted hexagonal (negative curvature) lipid phase by DSC or NMR, although this method does not really distinguish between changes in the bilayer elasticity or changes in the spontaneous curvature.

Many experimental studies exist

Axial diffusion is diffusion of lipid molecules within the same side of bilayer membrane. This kind of diffusion of lipids is very common and its rate is really high. There is another kind of diffusion of lipid molecules called "Lateral Diffusion" in which lipid molecules of the bilayer membrane of one side diffuses to the other side of lipid bilayer although this diffusion rate is quite slow

Usually dilution of the GUVs is more than enough to remove the background fluorescence because the signal from the inside of the GUVs is so strong. Keeping the GUV in the focus is usually more of a problem, but I assume you already have a found a way to do that (micropipette aspiration or the like).

I am not sure how you would see localized leaking by any method, considering the time resolution of the microscope and the diffusion constant of a small molecule. Your fluorophore will likely diffuse too quickly to get a spatial distribution. The only thing usually observable from these experiments is the fluoresecence average from the entire vesicle.

An alternate way to do it may be to use a water senstive probe embedded in the membrane (laurdan or prodan). If water is leaving through weak spots in the membrane, the laurdan fluorescence changes due to the polarity shift in the membrane. This method has been adapted to a confocal set up with GUVs

Yes, Pavel has the right idea. A D-spacing measurement gives you the repeat distance of bilayers in a stack, either in a multilamellar vesicle, or a stack of bilayers on a silicon wafer, such as in our oriented samples. The D-spacing is the combined thickness of the lipid bilayer and the water between the bilayers. It does not give you the lipid thickness directly. For fully hydrated fluctuating lipids in the fluid phase, we obtain the x-ray diffuse scattering which provides us not only material properties, such as the bending modulus, but also the structure directly. We take the square root of this intensity (form factor) and then the Fourier Transform, while fitting to a model of the lipid bilayer, where the Gaussian components are allowed to move to fit the data through the Fourier transform. It sounds complicated and it is, but our data agree nicely with MD simulations in many cases, including with peptides in the bilayer. Please read some of our papers which are available here in Research Gate or on our WEBsite: http://www.cmu.edu/biolphys/jfstn. Thank you. Stephanie

Thank You for your responses. I understand all the doubts you have raised regarding the outcome from a simple approach and how there are too many variables that need to be taken care of. But assuming that channel densities, pumps etc stay unaffected in between the control and mutants , would an enhanced lysis suggest a difference in membrane composition? I guess in my head that is what i was imagining, since rectifying channels would always complicate the system, i was just hypothesizing differences in lipid constituents leading to a propensity for easy rupture..

Hi Christopher,

I think you can try live cell imaging using FM dyes available with life technologies, which will enable you to track complete endocytosis(http://products.invitrogen.com/ivgn/product/F34653). The beauty of FM dyes are they are not fluorescent outside the endosome and starts fluorescing as it is taken up by the cells thereby allowing you to track the complete endocytosis and also exocytosis. Once you track this you can add your dyed bacteria together with FM dye and look for co-localization of dye with your bacteria.

I am not sure if this helps you, but I thought this information would be relevant for you.

I would also start with DSC, although FTIR, 31P and 2H NMR (you will need to ensure your lipids are available 2H labelled for the latter) do offer an alternative if you have them available. One consideration however when working with lipid mixtures is that the phase transition may be relatively broad.

Thank you!

Hi Joseph,

I have done some indentations on lipid bilayers with AFM before. There are also a number of groups around the world that does this. You can find references of groups doing research in this area of "AFM on lipid bilayers" from the the papers suggested by Marina. There are a number of parameters you can extract from your force curves that can somehow help you ascertain if you are getting real data or an artefact. As Rhiannon already pointed out, there's the breakthrough force, bilayer thickness (considering the average thickness of a lipid bilayer and the depth of water between the bilayer and the substrate, you can tell if what you're getting makes physical sense) and the depth of indentation. Of course, it goes without saying you should have a good statistics to get a better interpretation of your data. If you have access to force volume or force mapping, you might want to do that. What is your lipid composition? Do your lipid bilayers exist in the gel state, fluid-disordered, liquid-ordered? Those three states exhibit different magnitudes of breakthrough forces. You might be able to find some studies where you can compare your data against. Anyway, It is imperative that you first calibrate your cantilever. When I did it before, I am always in the look-out of a contaminated tip (which could give you double breakthroughs though you could also see double breakthroughs depending on your composition) or blunt tips that does not give you breakthroughs at all. These two papers might be of help:

1. Butt H.J., Franz V. Rupture of molecular thin films observed in atomic force microscopy. I. Theory. Phys. Rev. E. 2002;66:031601.

2. Loi S., Sun G., Franz V., Butt H.J. Rupture of molecular thin films observed in atomic force microscopy. II. Experiment. Phys. Rev. E. 2002;66:031602.

Best,

Ruby