DNA is an acronym for deoxyribonucleic acid. DNA has three chemical groups: bases (designated as A, C, G and T), phosphates and sugars. It exists in combination with proteins in entities called chromosomes. Embedded in the DNA are genes, which code for proteins. It is the sequence of DNA in the genes that makes people unique. There are 46 chromosomes, 22 pairs (two identical copies per pair) of chromosomes plus the sex chromosomes, X and Y. Everyone has two copies of each gene. For the sex chromosomes, XY indicates a male and XX indicates a female.
DNA is a double-stranded molecule and as such is exploited for molecular genetic tests because one strand of the molecule dictates the other. By separating the double strand, one can replicate the DNA and thus from one DNA molecule strand make two, from two, four, etc. DNA is important in newborn screening because it can be used as a tool to rule-in disease; however, in almost all cases, it cannot be used to rule-out disease.
The Human Genome Project set out to sequence the entire genome (all of the bases in DNA) in a linear fashion, like reading the pages of a book. The project also sought to develop informatics, sequence other organisms, develop advanced molecular technologies, and consider the ethical/legal/social implications of genome research. Originally, the National Institutes of Health used public funds to complete this project. After several years, a company, Celera, provided competition, saying they could read all 3 billion base pairs of the human genome faster. Both projects completed at the same time and were reported in two different journals. The NIH project was reported in Nature and the Celera project was reported in Science in mid-February, 2001.
There are three main forms of inheritance: autosomal dominant, autosomal recessive, and X-linked recessive.
In this disease the body becomes unable to break down fat stores during periods of poor nutrition or fasting, such as overnight or illness. The disease is caused by deficiency of the enzyme, which breaks down these fatty acids to produce energy. If children with MCADD are undiagnosed and not properly fed, they can lapse into a coma and die or become brain damaged. Newborn screening successfully identifies these infants and after diagnostic testing, treatment consists of keeping the child fed using sugar and cornstarch. If they become ill, a trip to the emergency room is in order. With this treatment, children with MCADD can live a normal life.
To identify infants who are candidates for further work-up, our program uses tandem mass spectrometry (ms/ms) to examine the levels of a compound with 8 atoms of carbon (octanoylcarnitine or C8). DNA is useful as a second-tier test because it helps to rule-in the diagnosis as only 1 mutation causes the majority of the disease. C8 levels may be elevated for other reasons, leading to false positive results.
Newborns with an elevated level of C8 (>1.0) are sent to the DNA Laboratory for molecular testing. A LightCycler instrument amplifies the DNA from the bloodspot using a process called polymerase chain reaction (PCR); next, without further manipulation of the sample, the instrument checks for the presence or absence of the mutation. Results are obtained in an hour or less, allowing results to be communicated the same day.
The result appears as a peak on the output scan and depending on the location of the peak, we can easily tell if the sequence is normal or mutant. Two peaks indicate the baby has one copy of the mutation and may be a carrier or have a mutation somewhere else in the gene. The test detects the most common mutation and determines if there is an A or a G at the 985th base pair of the gene. Another name for this change is K304E, which describes the change in the protein. This molecular shorthand result is called a genotype; phenotype refers to the physical nature of the disease. If a baby becomes very sick, we would say the genotype resulted in a severe phenotype. Another mutation we examine is called Y42H; this genotype is thought to result in a milder phenotype.
Our testing model or algorithm originally used a C8 level of 0.3, which means any sample above that number was sent for DNA analysis. In the first 18 months of screening, of the 385,893 newborn specimens tested by ms/ms, 511 samples were tested for mutations. Of those, 15 were clinically confirmed after additional diagnostic workup, which includes more precise biochemical analysis. Less than half of the mutations possible in the confirmed group were found to be the common mutation, so we sought to further investigate by sequence analysis, which examines every base (A, C, G, T) in the gene. We found 9 new variants that were not reported in the literature and 5 mutations that had been reported previously. Eleven samples had two mutations after sequence analysis. Only 1 mutation was identified in 3 babies and we could not find any evidence of mutation in 2 babies. These latter two figures suggest that there is a deletion in the DNA that sequence analysis could miss, or that another biochemical defect is mimicking disease in these infants. One baby was of Korean descent and the DNA Laboratory identified 2 new mutations. MCADD is rare in Asian populations.
After a period of experience, we modified the test model or algorithm for MCADD in late 2004. Currently, only samples with a C8 level greater than 1.0 are sent for DNA analysis. This was done in order to reduce the number of false positive screens, which are cases when the lab report shows a positive screen but in reality the baby does not have disease. Subsequently fewer samples were received in the DNA Laboratory, but still 3 of the 8 confirmed cases did not have the common mutation and 1 baby had only one copy of the common mutation detected. In order to provide the best service, we are in the process of expanding the test to include sequence analysis for cases such as these. Our data continue to show that the major mutation accounts for a smaller proportion of the total mutations than is indicated in the literature.
Currently we perform DNA analysis for cystic fibrosis (CF), an autosomal recessive condition. Each year the DNA Laboratory receives approximately 13,000 cases after the first CF test, which is called immunoreactive trypsinogen or IRT. IRT is found in the blood in higher levels in individuals with cystic fibrosis because they tend to have problems with pancreatic function. IRT can also be elevated due to a stressful birth, prematurity, and other factors.
The protein that is defective in individuals with cystic fibrosis is called cystic fibrosis transmembrane conductance regulator. It does its work at the surface of the cell and is responsible for the maintenance of salt concentrations. Babies with CF feed well but do not grow because they have problems digesting food and getting proper nutrition. Children with CF tend to be small and thin. Another organ most notably affected is the lung. CF patients have trouble breathing due to the accumulation of thick mucus in the lung. They are prone to pneumonia and other infections. Males with CF are infertile due to the absence of the vas deferens. Females with CF are fertile and can have children. Note all children from these individuals are at least carriers of CF.
Screening identifies infants early and the benefits include more normal growth and presumably more improved lung function. Screen positive infants (high IRT and 1 or 2 mutations) are referred to CF Centers for sweat chloride testing, which determines the amount of chloride in the babys sweat. High chloride levels indicate cystic fibrosis.
We use the PCR process, but instead of looking for 1 mutation, we look for 32 different ones at the same time. Patients with cystic fibrosis most often have 2 of about 25 common mutations, but many patients have more rare mutations (>1,100 have been reported). The most common mutation is called deltaF508. Thus to look for all 32 mutations, we must multiplex the PCR, meaning we amplify many different parts of the gene at the same time. Using a combination of size and color that we apply to the DNA pieces, we can then interrogate 32 different mutation sites by a process called capillary electrophoresis. The instrument we use is called an ABI PRISM 3100 Genetic Analyzer. Each run on this instrument examines 16 samples simultaneously. Each test plate of samples has six runs (96 samples per plate). The output or profile shows a series of peaks. If a mutation is present, the peak is labeled in red. The bottom of each profile has standards to make sure the sample ran correctly.
Of the 13,553 babies we tested in 2004, we have identified and received a diagnosis for 39 of them. Most of these infants have two mutations, identified by the screening program. Some babies sweat positive with only one mutation, but were referred by the program. If a baby has no mutations identified and the IRT level is still high (>170 or in the top 0.2% of samples for that day), the infant is referred based on only the IRT result. Two of these infants have been identified and diagnosed with cystic fibrosis since testing began in 2002. To our knowledge, we have not missed any infants with cystic fibrosis. We have gotten some special requests for infants to be tested who are now living in New York, but whose newborn screen was carried out in another state.
We are developing tests for other diseases using similar technologies. Some of these, galactosemia, biotinidase, hemoglobinopathies, and phenylketonuria were on the newborn screening panel prior to the November 2004 expansion. Others including propionic acidemia, long chain acyl-CoA dehydrogenase deficiency, and very long chain acyl CoA-dehydrogenase deficiency are newer additions to the panel. Many of the other new conditions will require full gene sequence analysis because the diseases are rare, less well studied, and the mutations tend to be private or exist only within a given affected family.
When developing new tests one must consider the number of mutations that are known to cause disease, the frequency of a given set of mutations, the turnaround time (receipt of the sample to report of the results) required for successful clinical intervention, the number of samples to be tested, and cost. As has been shown, the DNA tests in our laboratory currently use three different technologies: the LightCycler, fragment analysis (kit), and sequence analysis. It would be beneficial to consolidate the tests onto one platform, reduce the number of false positives (e.g. one mutation as in a carrier identified because of the high primary analyte, but no disease), and have automation at hand.
Thus, we have begun to examine other types of automated and comprehensive testing. Chip analysis has been promising in preliminary studies. This type of analysis would lead to DNA becoming the primary screen. There are technical and ethical considerations to overcome, but this analysis would substantially reduce the false positives resulting in referrals for only truly affected infants. This would result in a more comprehensive follow-up effort that could also examine outcome data. As you will hear, much of follow-up is dedicated to the tracking of large numbers of infants to ensure they are in a medical home.
Microarrays come in several different forms. For the purposes discussed here, it is a piece of glass approximately the size of a thumbnail. Onto this glass is placed a series of small pieces of DNA called oligonucleotides or probes. There are more than 10,000 of these probes on each chip. The process works by taking DNA from a bloodspot and cutting it into a series of pieces. The ends of these pieces are modified so they are all the same. Then the DNA is amplified using PCR as discussed above and this amplified DNA is loaded onto a cartridge holding the chip. The DNA washes over all the probes and by a process called hybridization, it sticks when all of the base pairs (between the probes on the chip and the DNA you physically add) are complementary. The chip is then read by software, which converts the data to a form that is understandable. Thus in one experiment, many pieces of data are obtained simultaneously.
Some companies whose technology would be amenable to newborn screening include: Affymetrix (described above), Illumina, Perlegen, and Nanogen. Other companies such as Agilent and Nanofluidics are working on lab-on-a-chip designs that include the delivery and maneuvering of very small volumes of liquid, far less than a small droplet. The goal of these tools is to add the bloodspot onto the cartridge and have the DNA extraction step and all testing be automated at once. One can even imagine point-of-care testing, where this is done in the hospital. In time, some or all of these tools will become commonplace and DNA may likely become the primary analyte studied in newborn screening laboratories.