Thursday, January 26, 2012

ACGT, or genomes & language & coding & clues

When I was a child, my dad worked for IBM in Essex Junction, Vermont. Of the several routes to the plant, which covered (and still does) acres and acres of land, my favorite ran alongside the bending path of the Winooski River until a left-hand turn by the large black sign brought us in, past a playground that might have been for IBM families (or not) and into the seemingly identical series of smaller streets among the beige brick and steel buildings. IBM was a gated city within a small Vermont town that had a single major intersection to its name.

The playground feels important because that was the fun part of IBM Family Day. Once a year, the company allowed us past security into the linoleum-tiled hallways to see the computer through a bank of windows designed for that purpose. THE computer, as there was only one and it occupied the perimeter of an enormous room that no one could enter unless they wore certain shoes and maybe hairnets. Most of what I saw at home was residue from the computer as Dad brought back some of the vanilla-colored punch cards that either drove its operation or were the output of its operation and we kept them by the phone to take messages upon. The cards had patterns made of small rectangular holes, clustered in various areas like a non-infectious disease.

The computer looked much as one might expect having watched many episodes of Star Trek: flashing lights and dials and pieces that looked like the wheels of giant tape recorders or film projectors spinning black shiny material clockwise and counter-clockwise. People who looked either vaguely scientific or somewhat business-like monitored what the computer did, which was buzz and spin and light up and spit out punch cards. Also reams of special paper connected continuously at the seams with perforated lines and perforated edges that attached marginal strips with evenly spaced holes to guide the wheels that turned the paper through the computer and its associated machines.

None of us really cared that much about seeing the computer, mainly because they wouldn't let us into the room and push the buttons ourselves, but we made an effort to be impressed because my dad found it so exciting to show it to us.

The laptop I'm typing these words upon probably has more functionality than that room full of metal and tape and paper and blinking lights. And the smart phone I hold in my hand has a significant fraction of the power of this laptop. The future comes toward us in ever smaller packages, the workings of the machines hidden from us by layers of software and shiny protective cases.

In mid-January, we made the drive to Baltimore to see Dr. Hamosh, the Hopkins geneticist who will guide us through the process of genome sequencing for Robert. So we made the familiar vehicular descent into the city on I-95 via that series of Jetsons-style elevated ramps above the marshlands of Baltimore Harbor, swooping past the clustered skyscrapers and into the heart of the medical campus, finding the right turns to the proper garage. The pediatric genetics clinic was accessed through the garage, through a series of doors equipped with big silver buttons for the handicapped, doors that felt like a series of airlocks into a space shuttle, when, actually, we found ourselves in a spacious waiting room the chief aspect of which was lack of windows.

We were in a basement, which seemed quite apt as, at this point, we don't know where the hell we are with diagnosis to begin with. Return to start, do not pass Go, do not collect $200.

Dr Hamosh is a brilliant researcher, scientist, doctor, whom I genuinely wished my daughter could meet, but Edith insisted that she had a mid-term exam and looked at me cross-wise when I tried to explain to her that I had infinite powers with the school to excuse her from such clap-trap so she could have a tutorial in genetics research and applications. In short, Edith did not believe me, although I believe my powers to be actual, probable superpowers of parental insistence and persuasion.

While I have long resided in the land of disability and neuro-genetics, a land governed by the simple tenet that little is really known and cause and effect are, ultimately, suspect, I admit to having certain simplistic understandings of what a genome is and what sequencing a genome involves.

Genes march two-by-two in chromosomal pairs like a biochemical Noah's Ark, right? Mendelian genetic theory, the monk and his pea flowers, in which floral characteristics obey some assuredly Darwinian logic in which dominant genes overpower the effects of recessive genes. Yes, sure, but also, not quite entirely.

All of the above, Mendel's tiny Xs (and in males, one lone Y) and their possible whole duplications and deletions are what can be seen under a high-powered microscope. We can now also perform DNA analysis with microarray, which allows us to see regions on each chromosome that might be variants, defects, duplications and deletions. Robert has had both of these analyses and he has no wholesale replication or absence of his various Xs, and no obvious defective regions (or add-ons) in the microarray analysis. One of my cousin's children was successfully diagnosed using basic DNA analysis, and another cousin's child was diagnosed using microarray.

Which, for Robert, leaves genome sequencing, an analysis of the four biochemical building blocks of each gene along each chromosome: adenine, cytosine, guanine, thymine, or ACGT. The patterns in which these letters are found can be interpreted to detect defects, deletions, additions, reversals, substitutions, repetitions, any of which scenarios may cause a shift, small or large, in the signals that initiate, maintain, or regulate biochemical processes. And, in turn, these biochemical processes as they build one upon the other, allow me to move my muscles, think, and type on this computer, among other tasks.

Genetic defects can be inherited, as in the Mendelian theory we all learned in high school biology, in which chromosomal replication creates gross (or subtle) problems during the earliest stages of reproduction and cell division. A fetus receives two, say, chromosomes 19, one from each parent, but each contains a recessive gene for a certain disorder not present in either parent as each parent's health is protected by a "good" X chromosome that corrects and takes over the directions issued by the faulty member of the pair. Chromosomes have co-dependent relationships, covering for each partner's failings.

Genetic defects can also occur spontaneously during the course of gamete replication in either the male or the female, or they can occur spontaneously during initial recombination as the gametes merge and reproduce. Some researchers even think that defects can occur over time in certain genes and the function of certain organelles as the messenger RNA makes errors transcribing from nDNA or mtDNA code. Remember, the mitochondria, which produce all of an organism's energy, have their own strand of DNA.

Time for an analogy. The best one I have is with computers. Gross chromosomal defects are like missing an essential program for your computer--or having duplicates running at once interfering with each other's signals and commands. All of a sudden, when you want to convert that Word document to PDF, Adobe Acrobat fails to comply. An error message results. No more PDFs for you.

Defects that can be determined through microarray analysis are like examining the command prompts that compose a computer's underlying operating system--DOS (disk operating system) ran (or still runs) PCs, and iOS runs Macs. On a PC, you can usually locate the program that reveals the DOS window and glance at the command prompts, issued in a string down the screen, that are causing the machine to function or malfunction. One of these coding languages was called ASCII, but I believe other languages are in use now.

Underneath the operating system is the actual computer code. All computer code is ultimately composed of the numerals 0 and 1, arranged in infinitely complex patterns, all of which make perfect sense to the computer, unless, somewhere, an incorrect string of them is embedded, or a programmer skipped a 1 somewhere or accidentally inserted an extra 0--for example. I'm sure I have some details wrong, but bear with me, analogies may contain imperfections, or they wouldn't be analogies, they would be some sort of higher level truth.

The future is handed to us in smaller and smaller packages. Some people say that good things come in small packages.

I say all of the above, all of it, to prepare you for what comes next.

Robert's blood was drawn without complaint from him by a phlebotomist, her office swelling with the strains of gospel music emanating from a small radio, racks of glass vials with color-coded rubber stoppers on shelves. Robert's caps were purple, as we those of mine and Roger. Our blood is being sent to Baylor University this week, where Dr. Hyland's lab (the neurotransmitter expert whose tests blew up the little shack constructed of diagnostic hypothesis we'd inhabited for many years) will perform whole exome sequencing on Robert's genome.

Each chromosome is composed of exons and introns, exons being the known coding area of the gene, and introns still tenuously referred to as non-coding regions of the gene. There are also transitional areas between exons and introns, as well as regulatory zones on each end of the chromosome.

Here is where I remembered Mrs. Lewis, as we settled into our seats for AP Chemistry in high school, telling us that everything we'd learned about electrons and molecules in regular Chemistry was taught to us deliberately wrong. We told you half-truths so you might understand.

Or, we might say, each set of scientific theories is like tract homes rising from a plain of newly cleared land, only to meet the fate of rapidly constructed living space--leaky roofs, badly sealed windows. Sometimes the contractors go back and try to repair their mistakes, sometimes they move on to the next project.

You see, introns are now suspected of containing genetic directives, as is the transitional area. Once, geneticists called introns "junk" DNA. I have a chromosomal variant on one of my chromosomes 20--a little "arm" that projects from a non-coding region, discovered when amniocentesis revealed that Edith had it. I was told that no one thinks this odd little piece of material does anything, or at least causes any disease, maybe, probably. I am the canary in the coal mine who continues to function much as a typical human being does. But who really knows what that mysterious arm does?

Regardless, we will not sequence Robert's introns. Why? Exome analysis will yield thousands of variants in the ACGT sequences in anyone. A computer will rapidly compile the hundreds of thousands of protein "letters" in Robert's exons and then do a sort of some kind by variety of error/defect. Then actual people need to sit down and pore over this data, not simply read it, but interpret it. Like the 0 and 1 sequencing of computer code, ACGT is a language, and all languages may appear at first to tell us absolute and blunt things, but any writer knows what substituting a verb can do to the meaning of a sentence, and any copy editor lingering over the best arrangement of clauses and phrases, the placement of commas (which are like punctuational variants), any copy editor knows just how imprecise and how variable language can be.

Yes, it all seems simple. Newspapers tell us that researchers have sequenced the human genome. At last! A template to which to compare Robert. But first, his little coding blips need to be compared to the letters in the same place on my genome and Roger's genome--if we have these blips, how do we interpret whether or not this variant causes or will cause disease?

And, I hate to tell you, but human beings are works in progress. Just because we think we know what the human genome is now means nothing. In 200 years, as DNA and RNA and gametes and the human sex drive do their day-to-day jobs, the genome will evolve to look somewhat different than it does now. That's right: any variation my family line encodes may simply be the next itty-bitty step in natural selection. And natural selection steps both forward and back--nature a vast experimenter, much like a three-year-old deciding it would be fun to dump nail polish into dish soap and see what happens. (If, in fact, as the genetic counselor cautiously probed my limits and biases, you do believe in evolution--cue the gospel music?)

Let me return you to Robert's neurotransmitter metabolites testing. In 2004, the first data we ever had on Robert's condition showed abnormalities!! At last, if not a diagnosis, a framework in which to place his movement disorder. In 2011, this test was returned to us saying that, in fact, everything looked normal. At our appointments with Drs. Gropman and Hamosh, we learned that, in fact, the two tests came back with exactly the SAME data, but now, seven years later, Dr Hyland knows that these results are common in children with neurological problems. No big deal.

And this is, really and finally, why we will not sequence Robert's introns. Researchers do not thoroughly understand intron coding, nor by what mechanisms intron genes are called into action, nor what those embedded ACGT sequences really mean. I could ask them to do it--it would cost more--but it might be a wasted effort. Whatever data we collected, whatever catalogue of defects we thought we had compiled might, in the end, be skewed as researchers, like developers in the heyday of the real estate and mortgage markets speculated only to be proven wrong in the end. And we'd have to do it over again.

I thought about all of this complexity this morning as I confronted my fears about what I was beginning to assume was a major plumbing problem behind the (gulp) drywall of our family room. I heard a dripping, a loud dripping, so loud it sounded like popping, just after the shower began running, just after water ran in the sink, just after the tub drained. It didn't happen just when the water ran, it didn't happen at all when no one was drawing water--water had to be draining, which eliminated water intake valves and pipes, and pointed to the PVC drain pipes, the main drain stack. But, maddeningly, I could find no evidence of water leaking anywhere: no damp spots, no swelling between the boards of the hardwood floors.

I envisioned Tom, our plumber, cutting open the drywall and imagined the expense not just of the plumbing repair, but the plastering, the painting, the reconstruction.

I called Tom, who is, frankly, a plumbing genius. He said, it could just be expansion and contraction of the pipes as hot water runs through and drains. Fill the sink with cold water, he said, and the tub. Then let it drain. Call me back if you still hear that dripping.

I didn't have to call him back.

Shit--how one thing can seem like another, how we'll talk ourselves into anything. How confusing clinical diagnosis is without data and knowledge applied. How I was sure Robert had mitochondrial disease because the description matched so closely what my observations had been all these years. How convinced many experts were as well. How I tried to talk myself into having early signs of multiple sclerosis while Dr Gropman enumerated them for me as adult onset MS happens often with parents of children with Leigh's syndrome. How the medical data is edging away from that mitochondrial diagnosis, bit by bit. How a couple of weeks ago, as Dr Gropman told us about tyrosine hydroxylase deficiency, a defect just available in the last six months as a genetic test. How I looked it up on the web and, open-mouthed, thought, that's exactly it!

My comfort here is knowing that six or seven or thirteen years ago, when doctors described a new theory, even I could see the pieces that didn't fit. As my observations and the clinical diagnostic parameters of new syndromes are fitting more and more closely, I know we are getting somewhere.

In about 90 days, we'll get our first report on Robert's sequencing: a list of anything that is known or highly suspected of causing disease. About four weeks after that, we'll receive a catalogue of all other defects that appear to actually disrupt function within a gene. That we can save with Robert's medical records to keep tabs on new research. Someone, someday, will decide to investigate many of those. And we can be connected with them. Maybe we'll get the real McCoy--a hard and fast diagnosis. Maybe we'll get more clues. We'll rule out TH deficiency, or will settle on it. And all of it will point more and more to possible affected areas of the genome and our understanding of what IT is will come closer and closer to knowledge and clarity.

And we'll continue to be prepared as the future comes to us bearing gifts wrapped in smaller and smaller packages--maybe tied with ribbon this time.

I feel comfortable in Baltimore now because I can find my way out through its nexus of one-way streets. After the appointment in genetics, we drove to Fells Point to have a very late lunch. Fells Point and its cobblestones--"how everything is waiting to be found," I though idly and nonsensically while I stared out the restaurant window at the 18th and 19th century buildings still standing in the shadows of 21st century construction a mile or two away. How the structure of a city still waits in place for us at the bottom of a hill, near the harbor, which smells like the not-so-distant sea.

No comments: