lp7 List of all known printed criticisms of the AI Hypothesis and their full rebuttal.


Sir Bernard Katz, Nobel Laureate wrote in his booklet: "Nerve Muscle and Synapse" (McGraw Hill, 1966): If cell K+ is selectively adsorbed, its mobility should be slower than in free solution. Experiments of Hodgkin and Keynes shows K+ mobility in isolated squid axon close to that in free solution (Hodgkin and Keynes, J. Physiol. (London) 119: 513, 1953). Similarly Kushmerick and Podolsky (Science 166: 1297, 1969) demonstrated K+ mobility in short frog muscle segment close to that in free solution. In response, Ling and Ochsenfeld (Science 181:78, 1973) showed that similar high K+ mobility in cell cytoplasm cited above can be reproduced at will if frog muscle cells were deliberately killed by prior exposure to metabolic poisons (or otherwise injured as in the region of the muscle cell close to the two cut ends of a short muscle segments used by Kushmerick and Podolsky in their studies). K+ mobility in healthy frog muscle cytoplasm is only 1/8 of that in free solution from 72 sets of experiments Ling and Ochsenfeld performed.

(See also Ling . Intern. Rev. Cytol. 26:1, 1969 for theoretical reason that while adsorption may slow down the rate of diffusion of K+ in frog muscle cytoplasm as Ling and Ochsenfeld has clearly demonstrated and discussed above, adsorption does not necessarily slow down the mobility of K+ or other ions. Thus, experimental observation showed acceleration of the mobility of adsorbed ion as on the surface of glass, for example).

Kushmerick later suggested an electrical potential field across the cut end of the muscle might spuriously affect the rate of K+ diffusion Ling and Ochsenfeld obtained ( Trends in Biochemical Sciences (TIBS) 3, Sept., N210). In 1974, my former gradual student, Chris Miller as a part of the reasons he offered for his rejection of the AI Hypothesis, raised a similar criticism (Chris Miller, Ph.D. Thesis, University of Pennsylvania, p.60). This criticism was fully and completely answered both on theoretical grounds and by experimental testing (Ling , TIBS, 4: N134, 1979).

In theory, an electric field effect suggested by Kushmerick and by Miller is not anticipated. The Law of Macroscopic Electroneutrality forbids K+ to travel alone in or out of the cut muscle cells in measurable quantity. It can only move in substantial quantities in or out of the cell (1) by exchange with one or more cation(s) traveling in the opposite direction, carrying a virtually identical amount of electric charges , or (2) accompanied by one or more kind of negatively charged ion(s) carrying a virtually identical amount of negative electric charges. For this reason, the overall phenomenon involved in the outward movement of labeled K+ movement in Ling and Ochsenfeld's study is an electrically neutral affair. As such, it is indifferent to the electrical potential field that may well exist across the cut end of the muscle. Experimental results described in the above-cited reference of Ling, 1979 further confirmed our theoretical anticipation. No influence of electrical potential field on the measured intracellular ion mobility could be observed.


In 1973 I reported experimental evidence affirming that in agreement with the AI Hypothesis, it is water polarized in multilayers at the cell surface which gives rise to its observed markedly higher permeability to water than to solutes (a phenomenon named by van't Hoff "semipermeability").(Biophys. J. 13: 807,1973). Later McElhaney (ibid 15, 777, 1975) contended as the title of his paper suggests: "Membrane lipid not polarized water responsible for the semipermeability properties in living cells". Biophysical Journal editor at the time, Frederick Dodge, refused to allow me to rebut at a length adequate to answer all the many attacks made in that same journal. In arguing my case, I lost much time to no effect. In the end, my rebuttal was published in another journal four years later (Physiol. Chem. Physics, 9: 301, 1977).In this rebuttal it was shown that without exception, McElhaney's criticisms came from his misunderstanding of what he criticized, other errors in understanding and presenting others' work he cited and still other errors of his own making.


Hinke and others inserted a K+-ion-sensitive glass microelectrode into a variety of giant cells and observed K+- ion activity---which one may define as an effective concentration--- close to that in a dilute solution of a free solution containing a similar concentration of K+ ion ( Nature 184: 1257, 1959; J. Physiol. (London) 156: 314). However, I pointed out that the K+-ion-sensitive microelectrode used cannot monitor the K+ -ion activity in the whole cell---as the experiment was intended to measure---but only that in a microscopic film of water surrounding the ion-sensitive tip of the microelectrode inserted into the cell and that this part of the cytoplasm is inevitably traumatized by the very same intruding electrode (Ling, Nature 221:386, 1969).

As work of this type expanded, the K+-ion activity recorded began to show wide fluctuations, ranging from K+-ion activity that is only a small fraction of the average K+ concentration of the cell, to activity which far exceed the average K+ concentration. Such variations are themselves at odds with the basic tenet of the membrane-pump theory, which requires all intracellular K+ activity measured in all living cells to be the same and equal to the K+ activity of that of a free water solution containing the same concentration of K+ as that in cells.

In a detailed analysis of the whole picture, Ling showed that the wide spectrum of data reported can be neatly explained on the basis of the basic tenets of the association-induction hypothesis: (1) water existing in the normal and healthy state of polarized multilayers has reduced solvency for K+ and Na+ ions; (2) cytoplasmic proteins offer adsorption sites for the selective adsorption of K+ ion when the cytoplasm is healthy and uninjured; (3) cytoplasmic proteins lose the ability of selectively adsorbing K+ ion in rough proportion to the extent of damage the cytoplasmic protein suffers; (4) K+ released from injured cytoplasmic proteins may find its way into the free water film in contact with the electrode while the remaining water remains largely uninjured. (5) K+ released from injured cytoplasmic proteins may find its way into the free water in contact with the electrode while the remaining water is also depolarized.(1) and (2) in sturdy cells can explain the lower K+-ion activity observed than that predicted from the measured K+-ion concentration; (5) can explain the earlier reported data where the observed K+- activity equals the K+-ion concentration; (4) can explain the observed K+-ion activity exceeding the K+-ion concentration (for more details, see Ling, "In Search of the Physical Basis of Life" Plenum, 1984, pp.250-257).


The supporters of the membrane-pump theory argued that an enzyme (K+-, Na+-activated ATPase) isolated from living cells---which can catalyze the hydrolysis of ATP in the presence of appropriate concentrations of Na+ and K+ ions--- is in fact the postulated sodium pump. In support, Goldin & Tong ( J. Biol. Chem. 249: 5907, 1974), Hilden, Rhee & Hokin (J. Biol. Chem.249:7432, 1974). Racker & Fisher (Biochem. Biophys. Res. Commun. 67:1144, 1975) and others incorporated isolated ATPase into phospholipid vesicles and showed that more radioactively labeled Na+ ion remained in the vesicles if the (presumed) energy source, ATP, was added to the buffer containing the labeled Na+. The authors attempted to explain the wrong direction of the Na+ "pumped" on the postulation that the membrane vesicle was inside-out, so that instead of pumping Na+ out of the vesicle making its Na+ level lower as it should, the pump was making the intracellular Na+-ion concentration actually higher.

In a detailed analysis of all the published data then accessible to us on the subject, Ling and Negendank (Persp. Biol. Med., 23:215-239,1980, pp. 224-236) showed that it was highly improbable what Goldin, Hilden and others observed and reported was what they claimed that they had demonstrated.

Ling and Negendank pointed out that the controlling step determining the level of Na+ ion in the vesicles could not be the initial loading (and the postulated pumping during that process). Rather, it was the leakage from the vesicles--- which the authors had overlooked---when the vesicles were subsequently passing through (the labeled Na+-free ) buffer solution in the Sephadex column---a step necessary in order to separate the radioactive Na+ ion trapped in the vesicles from the radioactive Na+ in the loading solution in which the vesicles were suspended before being loaded onto the Sephadex column. In other words, what they demonstrated was not how ATP activated the pumping of Na+ into the vesicle during the loading step. Rather it was that the inclusion of ATP in the loading process somehow had slowed down the subsequent leakage of labeled Na+ from the vesicles. That leakage process is a simple physical dissipative process and has nothing to do with the postulated energy-consuming pumping. One should also not forget that the concept of ATP containing a package of "high energy" has long been disproved in the fifties and early sixties. ( See linked page lp6c for references on the subject.)

Ling and Negendank then pointed out how these observation on the so-called synthetic ion-transporting systems can be better understood in terms of a long-confirmed part of the subsidiary theory of ion permeability in the AI Hypothesis.

In addition, Ling and Negendank mentioned that since healthy Nature-made cytoplasm -freed, cell-membrane sacs fail to pump Na+- or K+-ion (see linked pagelp6a{2}),would it not be somewhat presumptions to claim that vesicles prepared by Golden, Hilden and others cited above--- highly skilled biochemists though they unquestionably are--- did better?


In the AI Hypothesis, ATP, the ultimate metabolic product of living cells, controls the level of K+ ion in living cells by adsorbing onto specific protein sites (cardinal sites) and in so doing maintains the suitable electron density ( or more precisely, the c-value) of beta- and gamma-carboxyl groups on which K+ ion is preferentially adsorbed. Accordingly, there should be a quantitative relationship between the equilibrium level of ATP and of K+ ion in living cells when the level of ATP was made to change by controlled action of metabolic poisons.

Rangachari et al (Biochim. Biophys. Acta 274: 462, 1972) published K+ vs. ATP data from the study of rat myometrium. They concluded that the predicted linear relationship "did not always hold". A careful examination of their data showed that their data fully confirmed the original prediction except a single experimental data point (of high ATP and low K+ concentration). And that this point was produced by cooling the rat myometrium to zero degree Centigrade. In answer, Ling pointed out (Ling, Physiol. Chem. Phys. 6:285, 1974) that this departure actually lent additional support for the AI Hypothesis. The predicted quantitative relationships between K+ concentration and ATP concentration in living cells is restricted to observations made at the same temperature. Reduction of the temperature of warm-blooded mammalian tissues to 00 C as Rangachari et al did, brought into operation another aspect of the AI Hypothesis.

That is, the adsorption of K+ is "cooperative" (see Ling , Fed. Proc. Symp. 25:958, 1966) As such the K+ / Na+ distribution in mammalian cells undergoes a temperature transition characteristic of cooperative states between the K+-adsorbing state at high temperature and Na+-adsorbing state at a temperature below the transition temperature ( without perturbing the cell ATP concentration). And this is what Rangachari et al's single departing point confirmed.

In summary, Rangachari et al's data not only did not refute the prediction of the AI Hypothesis , they confirmed at once two basic aspects of the theory. (For additional discussion on temperature transition, see: Ling "In Search of the Physical Basis of Life" (Plenum, 1984), pp. 208-225 ; Ling "A Revolution in the Physiology of the Living Cell" (Krieger, 1992, Chapter 7; pp.188-196, also pp.293-294)


My former graduate student Peggy Neville and her coauthors (Science 184:1072, 1974; see also Civan and Shporer, Bioch. Biophys Acta 343:399, 1974) showed that frog lens and muscle when killed by heating showed a shortening of the NMR relaxation times of their water protons, in apparent contradiction to an expected lengthening, if the short relaxation times of normal living cells like those studied were due to motional restriction of the bulk-phase water molecules.

It is true that in the Polarized Multilayer Theory of the living cells (as an integral part of the AI Hypothesis), motional restriction of the bulk-phase water is clearly predicted. But as the author of this theory, I myself have never argued that the observed shortening of NMR relaxation times was exclusively or almost exclusively due to the motional restriction of bulk-phase water. On the contrary, I have repeatedly cautioned against this exclusive interpretation--- even though this interpretation came originally from scientists who had first provided NMR evidence in support of my theory---and have further pointed out that other factors including the rapid exchange with a small fraction of tightly bound water on paramagnetic ions and on cell proteins might play significant roles also (Ling, Intern. J. Neuroscience 1:129, 1970, p.138-139; Ling, in "The Aqueous Cytoplasm, ed. A.D.Keith, Marcel Dekker, Inc., New York, 1979, p.50-51; Ling and Tucker, J. Nat. Cancer Inst. 64: 1199, 1980, p.1206).

In years following, an important role of proteins (Koenig and Schillinger, J. Biol. Chem. 244: 3283, 1969) and of paramagnetic ions (in particular, manganese and iron and to a much lesser extent copper) in determining the NMR relaxation times of diverse tissues have been fully established (Ling, Physiol. Chem. Phys and Med. NMR, 15: 511, 1983; Ling et al, ibid, 22:1, 1990).Other model oxygen-containing polmers like poly(ethylene glycol), which does not have direct influence on water proton relaxation as proteins do, but also polarize water in multilayers, do cause substantial reduction of water proton relaxation times, T1 and T2 (Ling and Murphy, Physiol. Chem. Phys. , 15:137, 1983).

In work yet to be published, my former summer student, Gerri Magavero and I have shown that heating frog muscle in a Ringer's solution to a temperature (35 to 40 degree centigrade) enough to reduce/destroy the polarized multilayer dynamic water structure as monitored by the equilibrium distribution of sucrose probe (Ling in "Thermobiology" ed.A.H. Rose, Academic Press, New York, 1967, p. 19, Figure 10) does indeed produce lengthening of the spin-lattice relaxation time (T1); heating to 50 degree and higher , however, causes sharp reduction of T1 as Neville and others have found. Similarly, heating to 60 degrees or higher temperatures of aqueous solutions of most isolated pure proteins studied (though not all) caused a similar sharp reduction of T1. While one should not place too much emphasis on the T1 lengthening at 35 to 40 degree temperature (yet)---which could be at least in part be due to an accompanying cell swelling and increase of cell water content, the clear demonstration of T1 shortening in model protein-water leaves no question that the earlier observed lack of T1 lengthening offers no evidence against the polarized multilayer theory, since T1 lengtheing would be inevitably camouflaged by the pronounced shortening due to the cell proteins and protein-paramagnetic ion complexes. For demonstration of T1 shortening in model systems where water is polarized by other oxygen-containing polymers (e.g., poly(ethylene glycol)), but without the complicating contributions from proteins, see Ling and Murphy, Physiol. Chem. Phys. 15:137, 1983.


My former graduate students, Palmer and Gulati (Science 194: 521, 1976) claimed that their findings on the concentrations of cell K+ ion in frog muscle supports the membrane-pump theory---an 180 degree about-face from the beautiful work and conclusions of Gulati a few years before which remained rock-solid and unchallenged then and now (Gulati et al, Biophys. J. 11:973, 1971; Gulati, Ann. N.Y. Acad. Sci. 204: 337, 1973; Reisin and Gulati , ibid. 204: 358, 1973). However, that is far from the case in the newer work allegedly supporting their resuscitated belief in the membrane-pump theory.

In a rebuttal, Ling showed that the criticism of the AI Hypothesis came partly from a misunderstanding---incredible as it is to me--- of what they criticized, and that the general equation for solute distribution of the AI Hypothesis explains quantitatively all data including theirs as well as new data of my own, while their membrane pump theory---which can at best explain a part of the data (Ling, Science 198: 1281, 1977)---is tenable only by deliberately closing one's mind to a world of immediately relevant evidence to the contrary. (For other relevant information on the membrane-pump theory, see linked page lp6a and linked page 6b)


Another former graduate student, Jeffrey Freedman demonstrated selective uptake of K+ by, and selective extrusion of Na+ from red blood cell "ghosts"---red cells from which a major part of the intracellular proteins, primarily hemoglobin, has been removed by hypotonic lysis. Freedman saw in his finding a refutation of the AI Hypothesis, according to which, both ions distribute asymmetrically as a result of direct or indirect interaction with intracellular proteins---which Freedman believed he had removed from the red cell ghosts.

In a series of papers, Ling and coworkers showed that, contrary to Freedman's assertion otherwise, the specific method used by Freedman to remove all or virtually all hemoglobin (and other intracellular proteins proteins) does not do so at all. Rather, it retains different (as much as 25% of the original) amount of hemoglobin in the cell, depending on the individual blood donor. However, this retention of a variable amount of hemoglobin in the re-sealed ghosts, which destroys the alleged evidence for the membrane pump theory, turned out to be a great asset for further study. It has provided a mthod to vary the hemoglobin content of the red cells without bringing in additional treatments.

Using Freedman's procedure rigorously, we were able to show that both the amount of K+ regained and Na+ extruded in the subsequent incubation of "resealed" ghosts in the presence of ATP quantitatively depends on the amount of residual proteins (mostly hemoglobin) in the ghosts in full agreement with the prediction of the association-induction hypothesis. (Ling and Balter. Physiol. Chem. Phys. 7: 529-531, 1975; Ling and Tucker, ibid 15:311, 1983; Ling. Zodda and Sellers, ibid.16:381, 1984).With complete or near complete removal of hemoglobin from the ghosts, there was neither demonstrable re-uptake of K+ in, nor extrusion of Na+ from the resealed ghosts--- also in full support of the AI Hypothesis.


A major piece of experimental evidence against the membrane pump theory was my demonstration in 1962 that even if one assumes that the (postulated) sodium pump operates at 100% efficiency, and that the cells spend all their available energy to pump sodium ion only (under certain well-defined conditions), the minimum energy need of the sodium pump would be from 15 to 30 times the maximum available energy (see linked page lp6a).

In the Research News Report section of the Science magazine Volume 192 of 1976, science reporter, Gina Kolata cited from Drs. B and A for what she called their "crucial experiments and calculations" which, she claimed, have provided strong evidence for the existence of pumps (Science, vol. 192, p. 1220, column 2, line 4 from bottom). More specifically she stated that "Ling's data are compatible with a much lower rate of sodium efflux from the cell than Ling estimated. They (Drs. A and B) report that Ling's analysis of his data led him to assume that sodium was being transported out of the muscle cells at least 20 times faster than the rate accepted by muscle physiologists." (Science 192: p. 1222, column 1, 1976). The bizarre and totally erroneous story underlying this report and the full and complete rebuttal to all meaningful questions raised is given under linked page lp6b.


Horowitz (a former postdoctoral student) and Paine injected melted 10-20% gelatin solution into salamander eggs. On cooling, the injected gelatin solidifies into a semisolid gel globule. By analyzing the K+ and Na+ concentration in the gelatin globule, the authors claimed that they have affirmed the existence of Na+-K+ pump in the egg cell membrane.

Their evidence was built on the finding that after radioactively labeled Na+ had attained equilibrium with labeled Na+ in the bathing fluid, the level of labeled Na+ in the water of the gelatin globule was still only 33% of that in the bathing medium (Biophysical J. 25:45, 1979; see also Horowitz et al, ibid 25:33, 1979). They concluded that there must be a sodium pump in the egg cell membrane. However, I disagree.

I pointed out in 1984 that their conclusion is unwarranted because Na+ is not the only cation present in the gelatin globule. Present also in the globule was K+--- at an even higher concentration. Although labeled Na+ has reached diffusion equilibrium between water in the globule and in the bathing medium, K+ was far from having reached diffusion equilibrium. It was shown on thermodynamic grounds that the non-equilibrium, high concentration of K+ kept the Na+ level low by essentially the same mechanism that the presence of impermeant ion in a dialysis sac, also keeps other permeant solutes carrying the same electric charge at equilibrium levels below that in the bathing medium---in the well-known phenomenon called "Donnan Equilibrium". Thus the low Na+ level in the gelatin globule offers no evidence for the existence of the postulated Na+-K+ pump (Ling: Physiol. Chem. Phys. & Med. NMR 16: 293, 1984).


In 1930 A.V. Hill demonstrated that urea distributes equally between water in frog leg muscle and water in the surrouding medium (Proc. Roy. Soc.(London) Ser.B. 106: 477). This discovery was confirmed by the demonstration of similar equal distribution of ethylene glycol between water in surrounding medium and in living red blood cells (MacLeod and Ponder, J. Physiol (London) 86: 147, 1935) and between water in surrouding medium and water in frog abdominal muscle cells ( Hunter and Parpart, J. Cell. Comp. Physiol., 12: 309, 1938). These confirmative findings lent support to Hill's claim that water in living cells is simply normal liquid water in agreement with the membrane-pump theory. At the time, opponents to the membrane-pump theory were caught without an adequate answer for this set of observations; the membrane-pump theory gained a decisive victory, strengthening the belief by many in the membrane(-pump) thoery.

With the introduction of the AI Hypothesis and its subsidiary Polarized Multilayer Theory of Cell Water (PM theory), the situation changed. Thus according to the PM theory, water in living cells assumes the dynamic structure of polarized multilayers, in consequence of interaction with (various) cell proteins existing in the fully-extended conformation--- with their backbone carbonyl oxygen (and imino groups) at suitable distance apart and able to interact with solvent water. As such, cell water exhibits solvency for different solutes according to their molecular size and their surface molecular structure, when compared to their solvency in normal liquid water.

In this PM theory, urea and ethylene glycol distribute equally between cell water and surrounding medium because urea and ethylene glycol are small and because they possess surface structures compatible with the surroudning cell-water structure. However, in the same polarized water, the theory also predicts that larger solutes like sucrose and (hydrated) sodium ions should be found at a much lower level--- as it is well-known to be the case in virtually all living cells.

The PM theory also predicts that linear polymers carrying oxygen atoms (with their lone-pair electrons) at suitable distances apart---like the fully-extended protein chains in the living cell just mentioned ---should also be able to modify the solvency of bulk-water like that seen in living cells. This prediction has also been fully confirmed. Poly(ethylene glycol), poly(vinylpyrrolidone), gelatin, and urea-denatured proteins all satisfy the theoretical criteria of carrying properly-spaced oxygen atoms (with or without additional polar atoms). They are all capable of producing in bulk-phase water reduced solvency for sucrose, sodium ion etc. as seen in living cells, while at the same time, equal solvency for urea and ethylene glycol as Hill and others have demonstrated for living cells ( for details of quantitative theory and its confirmation, see Ling et al, Physiol. Chem. Phys. & Med. NMR 25: 177, 1993). (Note also that strictly speaming, the subject discussed under (11) was not a criticism of the AI Hypothesis, since the work of A.V. Hill was published long before the AI Hypothesis was introduced. It is nonetheless included to make sure that no major supporting evidence for the membrane(pump) hypothesis is left unanswered.