Oliver Smithies:

[00:00:00] — Book N’, March 16th, 2002, going to February 2003.  Starting the atrial natriuretic factor clone and targeting construct, again, the beginning on page 1, March 16th, Saturday.  No particular comments needed for the continuing of this type of construct, Thursday March 28th, page 13, 3’ to ANF targeting vector. [00:01:00]

Inserting the fragments into DupDel on page 19, Friday, April 5th, diagram showing what is going on there. [00:02:00]

Thinking about, and revising the construct on page 27, Friday April 15th, an ANF vector with hygromycin selection, in place of, or in addition to the neomycin selection.  And with the type of experiment needed to implement that, Thursday, April 25th, preparation of a hygro replacement fragment. [00:03:00]

Friday, the type of construct being considered again Friday April 26th, page 35, where the PGK promoter, phosphoglycerokinase, ubiquitously expressed promoter driving hygromycin, was being incorporated into the constructs.  So, [00:04:00] this is a little bit clearer diagram on page 37, ANF vector, hygrobeta, or hygro.

But looking at the hysteresis model, [00:05:00] beginning to understand that there’s a description of going back to the simplest model and realizing that the up and down event was extremely tightly-linked to the change of one variable.  The up is the first event, and the down is the second.  What was the sequence needed for flipping was being considered.  Thinking about [00:06:00] that this was related to the inflection point of the curve, Sunday, May 19th, continued to page 45.  It seems possible that the down transition is related to the inflection point on the phosphorylation versus stratograph, etc.  So theory, modeling, and plasmid construction going on at the same time, as Tuesday, May 21st, page 47, more constructs.  And whereas Sunday May 19th, modeling.   [00:07:00]

Wednesday June 12th, page 63.  This is following up on the work that we’d been doing with [00:08:00] Hyung-Suk Kim on the expression of different genes in the angiotensinogen-type experiments, which were between a lot of gender-related differences in gene expression, thinking about this problem, but never really being able to get at it properly.  A plethora of gender-related differences in gene expression.  It’s as I’ve said, and maybe earlier in talking about these various experiments, that I used to say, and still have said to Nobuyo that there’s more difference between males and females than there are between [00:09:00] mice and humans.  But — and it’s a longer history.  There is a longer history of male-female differences in mammals than there is of speciation of mammals, because males and females existed before mammalian evolution.  So that, the two genders are pseudo-species.  The time for evolution of male-female differences is much greater than the time of evolution of differences between any mammal, this idea being written down on page 63, Wednesday, June 12th. [00:10:00]

Monday, June 17th, anniversary of my sister’s birthday, and going by Buttercup which the five-year-old twins suggested that she ought to be called when she was born when we were five, reviewing the status of these constructs, for the nth time.  A lot of work, but not terribly productive in the end. [00:11:00] So you’d typically have page 83 with Wednesday, July 3rd being Amito CAN being looked at, a construct on the following page, and two, three days later, Saturday, July 6th, stellar flip-flopping again. [00:12:00] [00:13:00]

So, going back to page 95, Tuesday, July 15th, beginning to think about using what was called a Tie promoter-enhancer.  Tie is a factor where tyrosine kinase relate to immunoglobulin and epidermal cells.  I’m not quite exactly sure of what it was, but it was an [00:14:00] enhancer of interest to us.  It had been received from Thomas Sato by Suzanne Kirby.  And this construct was being made further on page 103, Wednesday, July 31st, continuing from M’95, in making this construct.  Drive the enhanced green fluorescent protein.  Tie-EGFP. [00:15:00]

Also, starting work on a GFP-gene driven by a HoxB4 cDNA, on page 109, Wednesday, August 7th, a variety of different constructs being made [00:16:00] still.  So, page 115, Tuesday, August 20th, assembled and checked as Hox IRES, Internal Ribosomal Entry Site, GFP. [00:17:00]

I went to Stockholm on September 6th, Friday, or shortly thereafter, and returned on Thursday, September 19th, presented work, a talk on beta-MHC vs. alpha-myosin heavy chain work, etc.  Let me do this again.

Page 130 summarizes rather important shift in my thinking, [00:18:00] which, page 131 of this notebook, which had considered a period of Friday, September 6th to Thursday, September 19th, is important insofar as my history is concerned, but not maybe as far as general science is concerned, but a change in the type of experiments which I was interested.  I went to Stockholm and presented the work at a symposium, nothing to do with the Nobel Prize.  Went to Stockholm and presented my work on the control of blood pressure, etc.  And, with a comment that, “Presented a [00:19:00] talk on the B-MHC vs. alpha-MHC, beta-myosin heavy chain vs. alpha myosin heavy chain, where a color slide on the left was how it ended.  The slide is missing.”  But during the symposium, Karl Tryggvason presented “Current views of the renal glomerular function,” and in which he reviewed current views that the slit diaphragm is the molecular discriminator that prevents albumin from being excreted in the urine.  And I asked him after the talk, “Why doesn’t it clog?”  And he answered, [00:20:00] — but before he could answer, I answered my own question with, “I know how: it’s the wrong way around.”  And I went on to quote Ogston and gel exclusion, the thought being that if the molecular sieve was at the slit diaphragm, that this would cause clogging of the basement membrane, that really the molecular sieve ought to be upstream close to where any accumulation of protein could be mixed by blood, by the passage of red cells, and over the next 12 days [00:21:00] or so, I refined this idea, and wrote a 500-word brief communication to Nature, of which the figure on the left on page 130 summarizes my thoughts, that the basement membrane was the region where the size discrimination occurred, and that it occurred because large molecules can permeate into a gel very poorly relative to small molecules, and so it will be a permeation type of gel, following Ogston’s equations.  And the diagram on page 130 summarizes this idea.  The paper [00:22:00] was not accepted.  But it did make me continue.

So, on the following page, Monday October 14th, page 133, which is about a month later, I comment that a short paper was sent to Nature as a brief communication was rejected, with a comment that they are not interested in hypotheses.  So, I need to work on a steady-state kinetics, etc.  And some comments on what I needed to be thinking about. [00:23:00]

On page 132 has the thought, their communication, listed there, or reproduced there, the one that was rejected.  So here I am saying what I need to do to improve the model.  I need to work on the steady-state kinetics.  And, realizing that it was very important to consider the change in GFR, glomerular [00:24:00] filtration rate, which depends also on the length of the slit diaphragm, as in minimal change nephrotic syndrome.  So, I began to make a model with Stella to try to understand what was happening in the kidney, on Saturday, November 16th, which is about a month later.  I commented that, “Considerable reading, literature review, and thinking results in five thoughts: that the transport of albumin, the cluster glomerulus, is determined by diffusion, and the [00:25:00] diffusion, in fact, accounts for most of the transfer of albumin, because the diffusion constant of albumin is not very much reduced in gels, and the dimensions are very small, and surprisingly, diffusion accounts for much more movement of albumin across the basement membrane than does the liquid flow, roughly 8:1 ratio.  And so that if one decreases the GFR, the concentration of albumin downstream will be altered, because the amount of albumin transported is very little affected by removing flow, but the transport of water, it will go very much [00:26:00] down, so that the concentration of albumin in the glomerular filtrate will increase, and can saturate the binding, or degradation sites in the tubule.

So, November 20th, a complete set of equations were worked out.   Managed to get all the elements of the data to be reproduced without much more complexity, but that it was hard work.  And, on the next page, page 139, Friday, January 24th, I describe the end result, and that on, I sent an extended version of the original Nature [00:27:00] paper for review for PNAS, and that in the end, it was accepted, and was published.  So, beginning of a new type of work, thinking about the kidney, and glomerular basement membrane permeation, starting then in 2003 and continuing to this day.

Back to constructs.  Back to the bench. [00:28:00] Monday, February 10th, page 141, making the HoxB4-driven GFP.  Return to constructs was interrupted, still now, with this new fascination of renal behavior, page 149, Saturday, February 15th.  Rankin and Gilmore data had shown that if you increase the concentration of albumin in the plasma, [00:29:00] that you would get albumin in the urine at a low concentration, and trying to see if my simulation would match the data.  And that did, quite well, as it says that using standard simulation and setting normal plasma albumin at 4.5g/100mL, and excretion at about 7.5mg per day, fine matches obtained without new assumptions.  So Rankin and Gilmore’s data, and the simulation data are both (inaudible). [00:30:00]

So the book ends still with construct-making, Friday February 28th, page 162, a repeat of the PCR for an Xho-Sbf fragment.  So ends Book N’.