A biochemical perspective on the use of tandem mass spectrometry
for newborn screening and clinical testing
Donald H. Chace*, Theodore A. Kalas
Pediatrix Analytical, 90 Emerson Lane, Bridgeville, PA 15017, USA
Received 15 November 2004; accepted 31 January 2005
Abstract
The first newborn screen was a clinical test to detect a disorder of the biochemistry of the amino acid, phenylalanine. This disorder, known as phenylketonuria, produces profound mental retardation if not detected and treated early in life. Early screening programs relied on inexpensive population screening techniques that have all but been replaced by more accurate analytical methods such as tandem mass spectrometry (MS/MS). MS/MS enables a multianalyte approach for detecting biochemical disorders such that a metabolic profile is obtained rather than a single analyte measurement. The metabolic profile has clearly shown improvements in the detection of diseases such as phenylketonuria and several new disorders arising from errors in fatty acid oxidation and organic acid metabolism. MS/MS is a powerful tool for accessing the metabolic status of a newborn and can detect both inborn metabolic errors as well as examine the effect of acquired diseases or pharmacologic intervention on intermediary metabolism. D 2005 Published by The Canadian Society of Clinical Chemists.
Keywords: Acylcarnitines; Amino acids; Tandem mass spectrometry; Newborn screening; Metabolism
Introduction
Blood was obtained at 28 h of age from two newborns, infant A and infant B. A report from the newborn screen of their dried blood samples included a quantitative pheny-lalanine (Phe) of 240 Amol/L and 360 Amol/L for patients A and B, respectively. These results were considered presumptive for phenylketonuria (PKU), as they were greater than a laboratory defined normal cutoff of 180 Amol/L. A second specimen was obtained from these infants at 8 days of age. The result of these 2nd specimens showed Phe concentrations of 720 Amol/L and 80 Amol/L for infants A and B, respectively. Infant A was seen by a metabolic specialist and sent to a PKU treatment center for further evaluation and testing, resulting in confirmation of PKU and initiation of dietary management. Infant B had a normal Phe concentration and was therefore considered normal with no further follow-up required. The results for Infants A and B were recorded as true positive and a false positive, respectively.
This scenario is often encountered in newborn screening laboratories where the numbers of true and false PKU results are affected by the established criteria for what is a considered a bnormalQ phenylalanine concentration alone. One question that should be asked is what produced the initial abnormal Phe in patient B? What could have been done differently to avoid reporting this patient as presumptive for PKU? Examination of the biochemistry of Phe can improve the understanding and characterization of alterations in the metabolism of Phe, leading to better use of clinical methods, improved reporting of results, and a more timely and cost-effective medical care.
This review will focus on how tandem mass spectrometry (MS/MS) can be used to improve clinical analysis for disorders such as PKU and how it also can be used to expand and improve newborn screening for other inherited metabolic problems. Unlike manuscripts pertaining to this MS/MS application that we and others have published [1,2], the focus of this review will be on the biochemistry and subsequent analysis of some key diseases that are a subset of the panel that can be screened using MS/MS. This review will also demonstrate that the choice of a metabolic screening approach that examines multiple metabolites is more likely to better characterize a particular metabolic disease or other iatrogenic influence such as total parenteral nutrition than a one-method, one-metabolite, one-disease approach. Three metabolic diseases will be described that represent disorders of amino acid, fatty acid, and organic acid metabolism. Emphasis will be placed on the laboratory result and how it can be interpreted clinically and better utilized in the process of diagnosing a disorder, integrating biochemical pathways where appropriate. Clinical situations typically encountered in a newborn screening laboratory will be used as illustrations.
Fundamentals of metabolic screening using mass spectrometry
Mass spectrometry (MS) is a technique that identifies and quantifies molecules based on molecular mass or weight. The elemental composition of a molecule will determine its mass while the arrangement and types of atoms will determine other physical and chemical characteristics, e.g., polarity, pKa, and volatility. Methods of introducing these molecules into a mass spectrometer are known as ionization techniques. In order to measure the mass of a molecule, the molecule must be present in a gas phase, so called vaporization, and it must possess a charge to be separated and detected. Since thousands of molecules may be present in complex mixtures and the intact ionized molecule (molecular ion) may share the same mass value as another, especially at low molecular weights (b1000), a secondary separation system is often used. The most common separation techniques are chromatographic systems such as gas
chromatography or liquid chromatography. Chromatographic separation occurs through identification of the molecular ion or fragment ions (pieces of the molecular ion that dissociate when subjected to sufficient energy and/or collision with other molecules). The use of two mass spectrometers in tandem (MS/MS) enables control of the formation of molecular and fragment ions. The instrument measures the mass of intact molecules in the first mass spectrometer, fragments them in a subsequent chamber known as a collision cell, and then measures the mass of these fragments in the second mass spectrometer. Fragments in the second mass spectrometer can be correlated with the intact molecules produced in the first mass spectrometer. This process enables unique mass spectrometry such as precursor ion scans and neutral loss scans. Since certain compound classes share common fragment ions of neutral fragment molecules, a special analysis can be set up to only detect precursor ions with a particular fragment. This results in an ability to measure only a particular chemical class or subset of molecules without detecting the hundreds of molecules that are not of interest. The net result is selectivity and speed while maintaining a comprehensive set of related metabolites in a 2-min assay. This process avoids time-consuming chromatographic separations that can take up to 30 min using a GC–or LC–MS system.
Results from a mass spectrometer are displayed in a mass spectrum (seeFig. 1) in which the horizontal axis represents the mass (m) to charge (z) ratio. Note that the axis values are essentially the same as the mass of measured compounds, since in this application the ions of interest have a charge of 1. The vertical axis represents the quantity of ions measured either in absolute terms (ion count, ion voltage) or relative to the ion with the highest intensity (relative intensity). Depending on the type of scan function (precursor ion scan or neutral loss scan), only acylcarnitines or amino acids are detected. Due to the speed of the scan processes, both molecular types can be obtained in one analysis, producing a comprehensive set of multiple mass spectra. With the addition of internal standards that are chemically the same as the molecules of interest but different by 3 or more mass units, the concentration of individual metabolites can be measured and quantified. A calculation of the concentrations of more than 65 metabolites and metabolite concentration ratios (molar ratios) is routinely performed in many laboratories analyzing both acylcarnitines and amino acids. (Note, for further discussion and information, see other reviews describing the use of MS in biomedical applications [3–5].)
Amino acid analysis by MS/MS
There are many useful analogies to describe how groups of atoms within molecules act as functional units and impart certain chemical and physical characteristics such as acidicity or hydrophobicity. These units within molecules, when subjected to high energy or other physical forces, will fragment in reproducible patterns. MS/MS can exploit these fragmentation patterns, creating special analyses that are selective for compound classes that share common function groups, i.e., alpha amino acids.
The MS/MS analysis of amino acids extracted from dried filter paper blood specimens was developed in 1990 [6,7]. Initial research focused on producing a simple mass spectrometric approach to measuring Phe, the key metabolite elevated in PKU. The original intention was to demonstrate two points: (1) the feasibility of an MS/MS approach to amino acid analysis, and (2) the investigation of the impact of a filter paper analysis for measurement of compounds versus more traditional liquid analyses.
The analysis of Phe extracted from a small disk of dried blood following a simple chemical modification, butyl esterification, and a subsequent selective analysis by MS/ MS using a neutral loss scan of 102 Da, led to a chance discovery. In the early profiles, peaks that were multiples of 14 and 16 in a full scan profile led to the recognition of other amino acids that shared a common chemical structure intrinsic to all alpha amino acids. Butyl esterified alpha amino acids, unlike gamma or delta amino acids, all fragment to produce a neutral molecule, butyl formic acid, and the remaining amino group containing the bRQ group. This is the common function group that all alpha amino acids share as described above, leading to the highly selective analysis for many amino acids we observe today. Further analytical and technical details regarding the manner in which MS/MS selectively detects alpha amino acids in a single analysis can be found in several good reviews and peer-reviewed articles [1,8,9].
Many, but not all alpha amino acids are easily detected using MS/MS; several important amino acids such as cysteine, homocysteine, lysine, and tryptophan require a special analysis or modification [10,11]. For example, amino acids with free thiol groups readily form disulfides which are doubly charged molecules or proteins with free thiol groups. Addition of a reducing agent is required to prevent these formations. These modifications would add substantially to the cost of obtaining a complete panel of amino acids and a separate analysis and, at this time, are entirely impractical for the purpose of newborn screening. Regarding other amino acids, gamma amino acids such as GABA (gamma amino butyric acid) require a separate scan function while AABA (alpha amino butyric acid) is detected in a NL 102 scan. In newborn screening for genetic disorders, the requirement for addition of AABA to the panel for detection of metabolic disease has not been demonstrated.
Amino acid profiles by MS/MS
All alpha amino acids fragment in a reproducible manner, producing a neutral molecule, formic acid. The standard MS/MS analysis requires derivatization to enhance ionization. The most common derivative is formed by esterification of the carboxylic function groups with butanol. These butyl ester derivatives of amino acids fragment as a butyl formate with a mass of 102 Da. Hence, the MS/MS analysis using an NL 102 Da scan will selectively detect amino acids as previously described. A typical NL 102 Da scan of amino acids extracted from a filter paper blood sample from a control patient is shown inFig. 1, top panel. Peaks representing various amino acids are labeled in bold text while the internal standards are italicized and underlined. A complete amino acid analysis requires the use of another type of MS/MS scan (data not shown) that selectively detects basic amino acids such as citrulline, ornithine, and arginine. This scan, also known as selected reaction monitoring (SRM), is useful in choosing a limited set of amino acids by selecting their specific mass values and fragment ions. Essentially, the difference between SRM and NL scans is that NL scans display an entire mass range while SRM scans choose only a few mass values. Furthermore, a special basic amino acid MRM scan enables optimization of MS analysis parameters and scan functions required for amino acids with differing fragmentation patterns of basic amino acids and MS tuning adjustments. What is ideal of MS/MS is that no secondary injections, separate sample analysis, or modifications in sample preparation are required. The analyses are all incorporated within the 2-min amino acid analysis.
NL 102 amino acid analysis acquires the data for phenylalanine and other important amino acids such as tyrosine, leucine, and methionine simultaneously [6,12,13]. This multiple metabolite approach offers the ability to interpret diseases in a way that is quite distinct from the single metabolite–single disease scenario that existed previously in traditional newborn screening and other laboratory tests. This approach is illustrated in the next section and is useful for understanding the complexity of the biochemistry of metabolic disease.
Newborn screening of amino acids—a biochemical approach
The biochemistry of amino acids is both complex and interdependent [14,15]. Amino acids have many functions in metabolism, serving as building blocks of peptides and proteins, precursors to hormone mediators and other functional molecules, and sources of energy production. A feature that distinguishes this group of metabolites from others such as fats and carbohydrates is the amino group. Many problems related to abnormal amino acid metabolism produce abnormal concentrations of ammonia, a result of increased turnover of amino acids for energy production or an indicator of alterations in urea cycle metabolism. From the newborn screening perspective, time is a critical component in the disease etiology. For most disorders, maternal metabolism insures near-normal concentrations of amino acids; after birth, however, deviations from endogenous metabolism are no longer kept in check. With time and other influences, such as diet, cell/protein turnover, and numerous other factors, both deviations from normal and compensatory mechanisms may occur. Many disorders have been viewed in children who exhibit serious symptoms of disease, in which the abnormal biochemistry is quite evident. In newborn screening, however, since many of these processes have only just begun and no outward medical problems have yet been presented, conditions in infants are all too easily affected by small changes in diet, collection time, and other such factors. Therefore, newborn screening will always remain a screening tool and never be 100% accurate. Newer technologies such as MS/ MS have impacted newborn screening in a way that has allowed it to more closely approach 100% disease detection through the use of more accurate technologies and multi-analyte approaches.
Phenylketonuria and multiple amino acid analysis
Phenylketonuria is categorized as a defect in the metabolism of phenylalanine caused by a deficiency of the enzyme phenylalanine hydroxylase [16–18] or its cofactor. Biochemical pathways that branch from phenylalanine are shown inFig. 2. The concentration of phenylalanine in blood is affected by the influx of phenylalanine into the blood-stream through dietary absorption, i.v. administration, and protein breakdown. The rate of elimination or turnover of phenylalanine is affected primarily by the enzyme phenylalanine hydroxylase (irreversible enzyme) and secondarily through phenylalanine transaminase (reversible enzyme). The products of phenylalanine that may be important in the assessment of phenylalanine metabolism include tyrosine, phenylpyruvic acid, phenylethalamine, phenylacetate, phenylacetylglutamine, and hydroxyphenylacetate.
The principle metabolites related to phenylalanine metabolism, measured by MS/MS, are phenylalanine and tyrosine. Other metabolites, such as the phenylketones listed above, are not detected because of the low sensitivity in the MS/MS methods utilized. Results of an MS/MS analysis of a normal newborn and a newborn with PKU using the NL 102 amino acid scan are shown in Fig. 1. Recent data from the analysis of 3000 control specimens collected from 4 models of tandem mass spectrometers from a single vendor (Applied Biosystems MDS-Sciex Models 300, 365, 2000, 3000) in 2 different labs over a period of several days show a mean Phe concentration of 49 Amol/L (0.8 mg/dL) and a mean tyrosine concentration of 64 Amol/L (1.2 mg/dL). These results compare well to published values of Phe in the range of 40–130 Amol/L (0.7–2.2 mg/dL) and Tyrosine in the range of 25–120 Amol/L (0.5–2.2 mg/dL). Cutoffs utilized in both labs were independently validated and were established to be nearly identical.
Traditionally (using BIA or fluorometry), only Phe is measured and used to detect PKU in newborn screening labs. Tandem mass spectrometry could theoretically be used in a similar manner (Phe only) by using the selected ion recording (SIR) scan mode. How would MS/MS fare compared to other technologies? In a retrospective study [19] comparing MS/MS analysis with fluorometry, the data show that the median concentration of Phe is lower, and in retrospect, had fewer false results when compared to fluorometry. In 96 cases out of 100, the concentration of Phe by MS/MS was within the normal range as compared to fluorometric analyses which showed an elevated Phe in all cases (above a threshold of 180 Amol/L [3 mg/dL]). MS/MS clearly improves the accuracy of the measurement of Phe alone.
However, the question or problem of further reducing false positives that are not due to false measurements of Phe remains. It is anticipated that these false results occur in less than 0.1% of analyses. In some screening laboratories, tyrosine is also measured in a separate, albeit similar, assay (with the exception of HPLC techniques adapted for newborn screening). Based on the metabolism of PKU, it is known that the enzyme deficiency of phenylalanine hydroxylase will result in decreased conversion of Phe to Tyr. Tyrosine would be obtained primarily from the diet, which presumably would be insufficient, resulting in a lower blood tyrosine level with time. It is known that Phe, in classical cases of PKU, will continue to rise with time. Therefore, the concentration of a metabolite that increases (Phe) divided by a metabolite that decreases (Tyr) would be a better indicator for detecting PKU. This concept of using molar ratios to improve analyses has been described in numerous publications [16,19–21]. In the retrospective study described above, 2 of 3 false positive results were resolved by utilizing the Phe/Tyr ratio. A significantly elevated Phe together with an elevated Phe/Tyr ratio is more indicative of PKU than any measurement of Phe alone[19].
Hyperamino acidemias
Is consideration of the ratio of Phe to Tyr alone sufficient to detect all cases of PKU and concurrently eliminate nearly all false positives? The answer is no. As will be described in more detail below, there are forms of supplemental intravenous amino acids given to infants that are very low in tyrosine (primarily due to poor solubility of Tyr at low pHs). This approach results in an elevated Phe without a concomitant increase in tyrosine and hence an abnormal Phe/Tyr ratio. There are other markers however that collectively can eliminate these false results. These markers are primarily other amino acids, which are measured in a full scan amino acid profile.
Other causes of elevated concentrations are due to immature enzyme systems in premature infants, non-genetic conditions that may affect liver function, iatrogenic causes such as total parenteral nutrition containing amino acids, or poor sample collection resulting in excess or clotted blood on the filter paper [22]. It is now possible to more reliably distinguish PKU from these other causes of elevated phenylalanine.
The most complete set of data from many important amino acids is achieved using a bfull scanQ of amino acids, i.e., a full range of mass values that correlate with over a dozen important amino acids. A generalized elevation of amino acids due to poor specimen collection characteristics such as double layering of blood can be recognized in a full scan of amino acids. Generally, all amino acids are mildly elevated in a profile such as this, in which all ratios such as Phe/Tyr are normal.
This same type of profile interpretation works with amino acid supplementation, but results can be much more variable due to the formula utilized. If a pharmacological mixture of amino acids is administered to a patient, its constitution and the quantity administered will determine whether the amino acid profile shows a moderate or severe generalized elevation of amino acids [1]. It is not unusual to observe an infant with a Phe concentration of 240 Amol/L (4mg/dL), while receiving total parenteral nutrition. PKU is ruled out by a normal Phe/Tyr ratio as well as other elevations in Leu, Ala, or methionine. Unfortunately, some common formulas or pharmacy-prepared formulas have an insufficiency of tyrosine (due to solubility problems in acidic preparations). In such cases, an elevated Phe/Tyr ratio is almost always observed. However, other amino acids are elevated and such ratios as Phe/Leu are normal, thus eliminating PKU as a possibility.
Of interest in this discussion of metabolism are tyrosinemias and their relationship to high Phe concentrations and generalized elevations of amino acids. A high concentration of tyrosine will affect the kinetics of phenylalanine hydroxylase (Fig. 2), limiting further production of tyrosine but resulting in a possible moderate increase in phenylalanine. A false positive result is more likely in laboratories that utilize methods that rely on the single measurement of Phe and is further exacerbated in situations where tyrosine can be cross reactive, e.g., fluorometry. Addition of a tyrosine assay, however, will assist in reducing false results, as an elevated tyrosine will be observed. Utilizing MS/MS to measure both compounds enables calculation of Phe/Tyr ratios in addition to individual concentration measurements. A normal to low ratio would be observed in cases of tyrosinemia when analyzed in the newborn period. In addition to an elevated Phe and Tyr in cases of tyrosinemia, other amino acids may be elevated, such as methionine, as occurs in Tyrosinemia Type I.
Regarding Tyrosinemia Type I, the elevation of Tyrosine is much reduced in the newborn period as compared to Types II and III and in fact may be normal. A separate assay that analyzes Succinylacetone is suggested for Tyrosinemia Type I screening.
More research is needed to investigate the significance of elevated amino acids in premature patients and others receiving amino acid supplementation. Do these high amino acid concentrations emulate or produce disease? Do they precipitate liver malfunction or other developmental abnormalities? Should supplemental amino acid formulas be monitored closely, more often, or individually adjusted to the metabolic needs of the patient? These are questions than can now be asked and will soon be answered by taking the first step towards examining the biochemical status of newborns.
Other metabolic disorders
In addition to neutral amino acids, detected using the NL 102 scan, an analysis of basic amino acids, ornithine, citrulline, and arginine can lead to the detection of urea cycle disorders. The basic amino acids are analyzed using a special scan based on the loss of the basic side chain in addition to the butyl formate [4].
Newborn screening of acylcarnitines, a biochemical approach
In addition to detecting several important amino acids, MS/MS can be utilized to measure free carnitine and acylcarnitines in the same derivatized extract from the dried blood specimen. Free carnitine is the bshuttleQ for transport of long-chain fatty acyl CoA compounds across the inner mitochondrial membrane where beta-oxidation can take place [1,15,23,24]. A diagram describing this process is presented in Fig. 3. Briefly, coenzyme A forms a highly energetic thioester bond with fatty acids in between the outer mitochondrial matrix. The fatty acid is transferred to carnitine, forming the fatty acyl ester of carnitine. This acylcarnitine is then transferred across the inner mitochondrial member and transferred back to coenzyme A, reforming the CoA thioester. This thioester proceeds through various oxidation steps resulting in shortening in the carbon length, two at a time, while producing acetyl CoA subunits. Defects in any of these steps can produce an accumulation of CoA thiosters, causing a parallel increase in acylcarnitines. It is the acylcarnitines described in this process that are measured by MS/MS, as they are analytically easier to detect using current ionization methods such as electrospray.
Acylcarnitine profiles by MS/MS
The MS/MS analysis of free carnitine and acylcarnitines extracted from dried filter paper blood specimens was introduced in the late 1980s [25,26]. Carnitine and its associated fatty acid esters have a quaternary ammonium group that is permanently charged. Analysis using techniques such as GC/MS is not possible without extensive sample preparation and compound modification. The focus of early research was to design a liquid chromatography approach. At the time, MS technology for the analysis of nonvolatile compounds using liquid matrices was advancing rapidly. New ionization techniques that utilize liquid matrices such as liquid secondary ionization MS, commonly referred to as Fast Atom Bombardment or Fast Ion Bombardment, enabled analysis of compounds such as acylcarnitines directly from a liquid matrix. Development of chromatographic systems to separate acylcarnitines from other components in a biological extract has been difficult. With the introduction of modern, affordable MS/MS systems that employed FAB or FIB ionization, the ability to separate acylcarnitines from other compounds was now possible. This technology required manual analysis of blood spot extracts, which limited high volume analysis to some degree. With a change of ionization techniques in the mid- 1990s to electrospray ionization, a high-throughput screening method was more readily achieved. The mass analysis techniques however have remained essentially the same for more than a dozen years [7,27].
Acylcarnitines and free carnitine share a common ion fragment that has a molecular mass of 85 Da. This enables a full scan profile of free carnitine and acylcarnitines (see Fig. 4, top panel) with the acyl group of 2 to 20 carbons in length. Like amino acids, acylcarnitines are butyl esterified to improve ionization efficiency when analyzed in a positive ionization model. It is especially helpful in improving the ionization efficiency of acylcarnitines that contain 2 carboxylic acid groups such as dicarboxylic acids. Further details regarding the manner in which MS/MS selectively detects acylcarnitines in a single analysis and how this analysis can be combined with the analysis of amino acids may be found in several good reviews and peer-review articles [4,28].
A potential biochemical energy crisis: MCAD deficiency and disorders of fatty acid oxidation
MCAD deficiency is categorized as a defect in the metabolism of medium chain length fatty acyl CoA compounds caused by a deficiency of the enzyme medium chain acyl CoA dehydrogenase [29–31]. Biochemical pathways that illustrate MCAD deficiency are shown in Fig. 3. As shown, MCAD is the first step in the oxidation of medium-chain length fatty acyl CoA substrates. The composition of the fatty acid side chain is believed to be a 6-carbon hexanoic acid, an 8-carbon octanoic acid, and a 10-carbon decanoic acid. Also included are these chain lengths with unsaturated double bonds, i.e., the 10-carbon, mono-unsaturated, decenoic acid. Note that the enzymes may also metabolize other acyl CoA compounds of different size, although less efficiently.
Considering this defect, it is not unexpected that the intramitochondrial concentration of medium-chain acylCoA compounds would increase substantially. These compounds presumably reform acylcarnitines via a carnitine palmitoyl transferase and are transported out of the mitrochondrial matrix. The acylcarnitines ultimately diffuse into the blood where they are transferred to the kidney or liver for excretion in urine or bile. In addition to acylcarnitine accumulation in blood, an expected increase in medium-chain fatty acids will occur.
Dysfunction of fat metabolism can cause a number of disorders, many of which can be fatal if not recognized and treated early [23,32,33]. Emphasis on disorders of fat metabolism that are detectable in the newborn period has grown over the past decade in recognition of the toxicity and the rapid death that may accompany such conditions [34].
From the analytical perspective, measurement of acylcarnitines rather than fatty acids is more efficient and sensitive. It is believed that acylcarnitines reflect the status of beta-oxidation within the mitochondrion. From a purely analytical point of view, there is little natural interference from the analysis of a fatty acid conjugate (an acylcarnitine) versus a free fatty acid analysis. Free fatty acids are present in the filter paper matrix and thus are common contaminants in the laboratory. Furthermore, fatty acids are most often and better analyzed via gas chromatography due to their physical nature.One interesting aspect of using a biomarker that is a conjugate is that a deficiency of the conjugate may result in a deficiency of the substrate [35,36]. In other words, formation of large concentrations of fatty acylcarnitines may deplete the concentration of unesterified carnitine. As the deficiency develops, so does the formation of acylcarnitines. Hence, detection of MCAD deficiency by an acylcarnitine may be impaired if it is in insufficient concentration. In a severe deficiency, the disease may be detected as a carnitine deficiency but of unknown origin. The challenging aspects of the assay are in interpretation of intermediate concentrations in which the free carnitine is not sufficiently low as to be considered deficient and the concentration of acylcarnitines is not sufficiently high to be demonstrative of disease.
As with PKU, however, the use of metabolic ratios may improve the detectability of this disease. A ratio of octanoylcarnitine to free carnitine or C8 to C2 or other acylcarnitine may be a better index of MCAD deficiency than the measurement of octanoylcarnitine alone, especially at low C8 concentrations [30,31]. Other ratios are also useful in distinguishing MCAD deficiency from other disorders; a discussion is provided below.
Regarding the relationship of carnitine and acylcarnitine measurement, newborn screening data have shown that the concentrations of carnitine and acylcarnitines are highest in the first few days of life and then decline over the subsequent days, so that by day 7, the concentrations of free and total carnitine are less than 50% of the peak concentration in the first 2 days of life [33]. This variability is one reason why different detection schemes are used for MCAD deficiency based on the age of the patient. In MCAD deficiency, as well as other fatty acid oxidation disorders, the accuracy of the test declines with age at collection and alternative methods of interpretation or other tests such as urine organic acids should be considered as complimentary tests.
The principle metabolites related to medium-chain fatty acid metabolism are C6, C8, C10, and C10:1, as measured using MS/MS [30,31,37,38]. Other metabolites, such as free carnitine, acetylcarnitine, and palmitoylcarnitine are also measured so that relative molar ratios can be determined. Results of an MS/MS analysis of a normal newborn and a newborn with MCAD deficiency using the precursor ion scan (Pre 85) acylcarnitine profile are shown in Fig. 4, bottom panel. Recent data from the analysis of 3000 control specimens collected from 4 types of mass spectrometers (models described previously) in 2 different labs over a period of several days for C6, C8, C10, and C10:1 reveal values of 0.08 Amol/L or less. This concentration is at the signal to noise limit of 3 times the baseline, which would produce an apparent concentration around 0.01–0.02 Amol/L. With this low level of detection, the cutoffs set by the laboratory vary widely. Published mean data show a suggested cutoff of not less than 0.4 Amol/L[31].
Related metabolic disorders
Some laboratories are now utilizing tandem MS in its most limited form, e.g., to measure only C8 and its internal standard rather than the panel of acylcarnitines and free carnitine. As with PKU detection and amino acid profiles, information provided by other acylcarnitines is helpful in leading to a more accurate laboratory interpretation, reducing false positives and false negatives. In the case of MADD versus MCAD deficiency, the concentration of C8 alone is insufficient to distinguish these diseases [24,42]. However, MADD, with its impairment of all dehydrogenases, produces inconsistent and unpredictable elevations in short-, medium-, and long-chain fatty acylCoA metabolism. This process sometimes results in elevations in short-chain acylcarnitines such as C4 and C5, medium-chain acylcarnitines (C6, C8, C10), and long-chain acylcarnitines (C12, C14, C16). What is interesting is that the most diagnostic metabolite for MADD may in fact be, C12, since it is metabolized by both medium- and long-chain acylCoA dehydrogenases. Therefore, in either VLCAD or MCAD, it may not be elevated because it can be metabolized by the unaffected enzyme. Valproate therapy appears to increase, in addition to C8, the concentration of C10 more than C10:1. It has been reported that the relative ratio of C8/C10 may be a useful diagnostic tool for distinguishing these two disorders[31].
It should be readily apparent by now that the use of complete profiles enables improved comprehensive detection of metabolic disease and metabolic assessment. The concept of analyzing a single metabolite with a technology that easily obtains many metabolites is inefficient and an example of improper application of resources. This will be further illustrated in the final example for organic acids, detection of propionic and methylmalonic acidemias.
Metabolism of metabolites—a tale of two organic acidemias
Are propionic acidemias and methylmalonic acidemias disorders of amino acid metabolism or are they disorders of organic acid metabolism? Deamination of amino acids produces organic acids, which are metabolized to a variety of substrates that can be used to generate energy or simply used as substrates for biosynthesis such as cholesterol (Fig. 5). A disorder in the metabolism of organic acids [43,44], whether derived from amino acids and other compound classes, is an organic acidemia. In fact, propionic acidemia or methylmalonic acidemia is a disorder of those organic acids that were originally derived from multiple amino acids, e.g., isoleucine and valine. Of course, organic acids are derived from many different substrates, and in the case of propionic acidemia or methylmalonic acidemias, the original substrates are primarily branched-chain amino acids. It is more appropriate however to describe these diseases as organic acidemias and not amino acidemias since the defect involves metabolism of specific compounds that are considered to be organic acids (albeit derived from amino acids, lipids, nucleic acids, cholesterol, etc.). This metabolism is illustrated in Fig. 6. Note that the CoA thioesters formed in the metabolism of organic acids enter into the beta oxidation cycle (part of which is illustrated in Fig. 3).
Acylcarnitine profiles by MS/MS
The MS/MS analysis of propionylcarnitine and other acylcarnitines is described above in the section for MCAD deficiency. One notable discussion regarding acylcarnitines not previously described involves those organic acylcarnitines that are derived from dicarboxylic acids. The carnitine ester of methylmalonic acid is detected along with other acylcarnitines of importance, e.g., propionylcarnitine, in the acylcarnitine profile. Analysis of dicarboxylic acids is most sensitive following esterification using butyl esters. An understanding of the fundamentals of ionization best explains why this is the case. Although the value of detecting methylmalonylcarnitine for detection of methylmalonic acidemias is questionable, it is certainly not the case for other dicarboxylic acids, namely, glutaric and malonic acids, where the dicarboxylic acids are reliable for disease detection.
Biochemical screening: propionic and methlylmalonic acidemias
Propionic and methylmalonic acidemias are categorized as defects in the metabolism of the organic acyl CoA compounds resulting from a deficiency of either Propionyl CoA Carboxylase (PA) or Methylmalonyl CoA Mutase (MMA). The characterization of these disorders is actually more complicated as shown in the schematic in Fig. 6. Metabolism of Propionyl CoA, Methylmalonyl CoA, and Succinyl CoA are reversible enzyme steps and, if an enzyme’s functions were impaired, both product and substrate metabolism would be altered in a manner quite different than that observed in other enzyme deficiencies. In addition, these enzymes, propionyl CoA carboxylase or methylmalonyl CoA mutase, require biotin and cobalamin, respectively, as cofactors for proper function. Defects in its function therefore could be due to problems in the enzyme structure, a deficiency of cobalamin, or some impairment of the interaction between the cofactor and enzyme. In MMA and PA, the key metabolite that is elevated is propionylcarnitine produced from propionyl CoA. However, each of these diseases should be discussed separately in terms of detection.
Propionic acidemia
The principle metabolite related to propionic acidemia[45] is propionylcarnitine as measured using MS/MS. Other metabolites, such as free carnitine, acetylcarnitine, and palmitoylcarnitine are also measured, so that relative molar ratios can be determined in a manner similar to that described previously for MCAD deficiency. Results of an MS/MS analysis of a normal newborn and a newborn with propionic acidemia using the Pre 85 acylcarnitine scan are shown in Fig. 6 in the top and middle panels. Recent data from the analysis of 3000 control specimens collected from 4 types of mass spectrometers in 2 different labs over a period of several days reveal a median C3 of 1.65 Amol/L or less. Published mean data show a suggested cutoff of C3 that is quite variable and is highly dependent upon ratios of C3 to C2 and C3 to C16. With this variability, there is the potential for a moderate degree of false positive or false negative results unless a sophisticated algorithm is utilized.
Generally, newborns with propionic acidemia have an unusually high concentration of C3. Further, the concentration of free carnitine may diminish rapidly in the first few days of life. Because this disorder can be quite severe, early detection and characterization are important since there is often little time for confirmation. The key consideration is the elimination of false or borderline results; it appears that an examination of other factors for C3 elevation is also important. These observations will be described following the next section on Methylmalonic acidemia.
Methylmalonic acidemia
As with propionic acidemia, the principle metabolite related to detection of methylmalonic acidemia [45] is propionylcarnitine, as measured using MS/MS. Other metabolites such as FC, C2, and C16 are important and in fact, it is quite difficult to distinguish a severe case of MMA from PA using MS/MS. As previously described, the median concentration of C3 in cases of PA is generally greater than that for MMA, but variation is great enough to make distinction of the two particular disorders poor. The main features of the acylcarnitine profile that make recognition of this group of disorders easier are the same, however, namely, an elevated C3, C3/C2 ratio and C3/C16 ratio. It might be suspected that a disorder of methylmalonic acid metabolism would also produce an elevation of methylmalonylcarnitine in an acylcarnitine profile generated by MS/MS. This assumption, however, has proven false and the use of methylmalonylcarnitine to assist in the detection of MMA in newborns is not reliable. Not all cases of MMA produce very high concentrations of C3 and, in fact, MMA is a loose grouping for a series of disorders that range from the severity of methylmalonyl CoA mutase enzyme dysfunction to a deficiency in the activity or supply of Cbl cofactors required for function of this enzyme. Specific defects include mutase + and (activity detected or not) or the absence of proper cofactor functioning due to a defect in cobalamin synthesis or vitamin B12. In some cases, a case of MMA in a newborn may be a temporary state due to a defect in vitamin B12 metabolism in the mother, causing an apparent metabolic defect in the newborn.
Other metabolic disorders
As in PKU and MCAD deficiency with their primary metabolites, Phe and C8, the primary metabolite of interest in propionic and methylmalonic acidemias, C3, may also be elevated in other diseases. For instance, elevated C3 concentrations are sometimes seen in beta-ketothiolase deficiency and other disorders associated with 3-hydroxyisovalerylcarnitine elevations, namely, 3-methylcrotonyl-CoA carboxylase (3-MCC) deficiency, 3-hydroxy-3-methylglutaryl-CoA (HMG) lyase deficiency, and 3-methylglutaconyl-CoA (3-MGA) hydratase deficiency. Furthermore, supplementation with carnitine can cause an elevation in C3 and other short chain acylcarnitines. When carnitine is administered, high concentrations of C3 and C2 are commonly observed. Presumably, C3 is elevated due to a generalized increase in C2, which causes impairment of the breakdown of other short-chain acylcarnitines. Although it seems reasonable to suggest that a dietary deficiency of vitamin B12 could also produce an elevation in C3, there is little evidence in this respect. The most severe vitamin B12 deficiency cases, as described previously, involve problems with absorption of B12 and its effect on the newborn. Few studies have examined the effect of dietary folate, B12, on fatty and amino acid metabolism [46,47].
Challenges in detecting PA and MMA
Interpretation of MS/MS data for PA and MMA can be challenging. In most positive cases, the increase in the concentration of the diagnostic metabolite, C3, is quite high[45]. In late onset or less severe cases, however, the concentration of C3 may only be moderately elevated, and in even rarer cases of some cobalamin disorders, it may be within the normal range for C3. Clearly, of the 3 diseases described here, detection of the metabolite C3 in PA and MMA is the most problematic from the false positive perspective. Throughout more than 10 years of data interpretation, our laboratory has used and tested numerous ratios and cutoffs for C3. In our experience, in addition to the concentration of C3, the ratio of C3/C2 has proven to be the most important diagnostic measure in newborns and older infants [45]. The ratio of C3/C16 is also helpful in newborns but declines in utility with age, in part because of the rapid decrease over time in long-chain acylcarnitines relative to short-chain acylcarnitines. A further understanding of metabolism and other metabolites may help to reduce the degree of false positive results.
Summary
There are many benefits to using a metabolic screening approach that measures multiple metabolites and produces a comprehensive metabolic profile over traditional one-method, one-metabolite, one-disease approaches. We have described the underlying biochemistry of certain inborn errors of metabolism to illustrate how altered metabolic states are expressed and subsequently detected by MS/MS. By providing improved detection of specific disorders of amino acid, organic acid, and fatty acid metabolism, and allowing for identification of iatrogenic and other influences on the metabolic status of a newborn, tandem mass spectrometry has allowed for substantial improvements in newborn screening and clinical testing.
References
[1] Chace DH, Kalas TA, Naylor EW. The application of tandem mass spectrometry to neonatal screening for inherited disorders ofintermediary metabolism. Annu Rev Genomics Hum Genet 2002; 3:17 – 45.
[2] Rashed MS, Bucknall MP, Little D, et al. Screening blood spots for inborn errors of metabolism by electrospray tandem mass spectrometry with a microplate batch process and a computer algorithm for automated flagging of abnormal profiles. Clin Chem 1997 (Jul);43(7):1129 – 41.
[3] Willoughby R, Sheehan E, Mitrovich S. A Global View of LC/MS. How to Solve Your Most Challenging Analytical Problems. 1st ed. Pittsburgh, PA7 Global View Publishing; 1998.
[4] Chace DH. Mass spectrometry in the clinical laboratory. Chem Rev 2001 (Feb);101(2):445 – 77.
[5] Desiderio DM. Mass Spectrometry: Clinical and Biomedical Applications. New York7 Plenum Press; 1992.
[6] Chace DH, Millington DS, Terada N, Kahler SG, Roe CR, Hofman LF. Rapid diagnosis of phenylketonuria by quantitative analysis for phenylalanine and tyrosine in neonatal blood spots by tandem mass spectrometry. Clin Chem 1993 (Jan);39(1):66 – 71.
[7] Millington D, Kodo N, Terada N, Roe D, Chace D. The analysis of diagnostic markers of genetic disorders in human blood and urine using tandem mass spectrometry with liquid secondary ion mass spectrometry. Int J Mass Spectrom Ion Processes 1991;111:211 – 28.
[8] Sweetman L. Newborn screening by tandem mass spectrometry:gaining experience. Clin Chem 2001 (Nov);47(11):1937 – 8.
[9] Leonard JV, Dezateux C. Screening for inherited metabolic disease in newborn infants using tandem mass spectrometry. BMJ 2002(Jan 5);324(7328):4 – 5.
[10] Magera MJ, Lacey JM, Casetta B, Rinaldo P. Method for the determination of total homocysteine in plasma and urine by stable isotope dilution and electrospray tandem mass spectrometry. Clin Chem 1999 (Sep);45(9):1517 – 22.
[11] Accinni R, Campolo J, Parolini M, et al. Newborn screening of homocystinuria: quantitative analysis of total homocyst(e)ine on dried blood spot by liquid chromatography with fluorimetric detection. J Chromatogr B, Analyt Technol Biomed Life Sci 2003 (Mar 5);785(2):219 – 26.
[12] Chace DH, Hillman SL, Millington DS, Kahler SG, Roe CR, Naylor EW. Rapid diagnosis of maple syrup urine disease in blood spots from newborns by tandem mass spectrometry. Clin Chem 1995 (Jan);41(1):62 – 8.
[13] Chace DH, Hillman SL, Millington DS, Kahler SG, Adam BW, Levy HL. Rapid diagnosis of homocystinuria and other hypermethioninemias from newborns’ blood spots by tandem mass spectrometry. Clin Chem 1996 (Mar);42(3):349 – 55.
[14] Christenson R, Azzazy H. Amino acids. In: Burtis CA, Ashwood ER, Tietz NW, editors. Tietz Textbook of Clinical Chemistry. 3rd ed. Philadelphia7 W.B. Saunders; 1999. p. 444 – 76.
[15] Scriver CR. The Metabolic and Molecular Bases of Inherited Disease. 7th ed. New York7 McGraw-Hill Health Professions Division;1995.
[16] Dougherty FE, Levy HL. Present newborn screening for phenylketonuria. Ment Retard Dev Disabil Res Rev 1999;5:144 – 9.
[17] Shintaku H. The metabolic basis of the hyperphenylalaninemias and tyrosinemia. Nippon Rinsho 1992 (Jul);50(7):1542 – 7.
[18] Phenylketonuria (PKU) screening and management : January 1980 through July 2000, plus selected earlier citations : 3,394 citations [computer program]. Version. Bethesda, Md. (8600 Rockville Pike, Bethesda 20894): U.S. Dept. of Health and Human Services Public Health Service National Institutes of Health National Library of Medicine Reference Section; 2000.
[19] Chace DH, Sherwin JE, Hillman SL, Lorey F, Cunningham GC. Use of phenylalanine-to-tyrosine ratio determined by tandem mass spectrometry to improve newborn screening for phenylketonuria of early discharge specimens collected in the first 24 hours. Clin Chem 1998 (Dec);44(12):2405 – 9.
[20] Schulze A, Kohlmueller D, Mayatepek E. Sensitivity of electrospraytandem mass spectrometry using the phenylalanine/tyrosine-ratio for differential diagnosis of hyperphenylalaninemia in neonates. Clin Chim Acta 1999 (May);283(1–2):15 – 20.
[21] Reilly AA, Bellisario R, Pass KA. Multivariate discrimination for phenylketonuria (PKU) and non-PKU hyperphenylalaninemia after analysis of newborns’ dried blood-spot specimens for six amino acids by ion-exchange chromatography. Clin Chem 1998 (Feb);44(2):317 – 26.
[22] Hannon WH, Boyle J, Davin B, Marsden A, McCabe ERB, Schwartz M. Blood collection on filter paper for neonatal screening programs. NCCLS Doc LA4-A3; 1997.
[23] Bennett MJ, Rinaldo P, Strauss AW. Inborn errors of mitochondrial fatty acid oxidation. Crit Rev Clin Lab Sci 2000 (Feb);37(1):1 – 44.
[24] Rinaldo P, Matern D, Bennett MJ. Fatty acid oxidation disorders. Annu Rev Physiol 2002;64:477 – 502.
[25] Norwood DL, Kodo N, Millington DS. Application of continuous-flow liquid chromatography/fast-atom bombardment mass spectrometry to the analysis of diagnostic acylcarnitines in human urine. Rapid Commun Mass Spectrom 1988;2(12):269 – 72.
[26] Millington DS, Norwood DL, Kodo N, Roe CR, Inoue F. Application of fast atom bombardment with tandem mass spectrometry and liquid chromatography/mass spectrometry to the analysis of acylcarnitines in human urine, blood, and tissue. Anal Biochem 1989;180(2):331 – 9.
[27] Millington DS, Kodo N, Norwood DL, Roe CR. Tandem mass spectrometry: a new method for acylcarnitine profiling with potential for neonatal screening for inborn errors of metabolism. J Inherit Metab Dis 1990;13(3):321 – 4.
[28] Carpenter KH, Wiley V. Application of tandem mass spectrometry to biochemical genetics and newborn screening. Clin Chim Acta 2002 (Aug);322(1–2):1 – 10.
[29] Rinaldo P, O’Shea JJ, Welch RD, Tanaka K. Diagnosis of medium chain acyl-CoA dehydrogenase deficiency by stable isotope dilution analysis of urinary acylglycines: retrospective and prospective studies, and comparison of its accuracy to acylcarnitine identification by FAB/ mass spectrometry. Prog Clin Biol Res 1990;321:411 – 8.
[30] Van Hove JL, Zhang W, Kahler SG, et al. Medium-chain acyl-CoA dehydrogenase (MCAD) deficiency: diagnosis by acylcarnitine analysis in blood. Am J Hum Genet 1993 (May);52(5):958 – 66.
[31] Chace DH, Hillman SL, Van Hove JL, Naylor EW. Rapid diagnosis of MCAD deficiency: quantitatively analysis of octanoylcarnitine and
other acylcarnitines in newborn blood spots by tandem mass spectrometry. Clin Chem 1997 (Nov);43(11):2106 – 13.
[32] Rinaldo P, Matern D. Disorders of fatty acid transport and mitochondrial oxidation: challenges and dilemmas of metabolic evaluation. Genet Med 2000 (Nov–Dec);2(6):338 – 44.
[33] Chace DH, DiPerna JC, Mitchell BL, Sgroi B, Hofman LF, Naylor EW. Electrospray tandem mass spectrometry for analysis of acylcarnitines in dried postmortem blood specimens collected at autopsy from infants with unexplained cause of death. Clin Chem 2001;47(7): 1166 – 82.
[34] Bennett MJ, Variend S, Pollitt RJ. Screening siblings for inborn errors of fatty acid metabolism in families with a history of sudden infant death [letter]. Lancet 1986;2(8521–22):1470.
[35] Scaglia F, Longo N. Primary and secondary alterations of neonatal carnitine metabolism. Semin Perinatol 1999 (Apr);23(2):152 – 61.
[36] Van Hove JLK. Carnitine in Inborn Errors of Metabolism, From Pathophysiology to Applications and Treatment. Durham, NC7 Department of Pediatrics, Duke University Medical Center; 1994.
[37] Bennett MJ, Ragni MC, Ostfeld RJ, Santer R, Schmidt-Sommerfeld E. Population screening for medium-chain acyl-CoA dehydrogenase deficiency: analysis of medium-chain fatty acids and acylglycines in blood spots. Ann Clin Biochem 1994;31(Pt 1):72 – 7.
[38] Clayton PT, Doig M, Ghafari S, et al. Screening for medium chain acyl-CoA dehydrogenase deficiency using electrospray ionsation tandem mass spectrometry. Arch Dis Child 1998 (Aug);79(2):109 – 15.
[39] Chace DH, Kalas TA, Naylor EW. Use of tandem mass spectrometry for multianalyte screening of dried blood specimens from newborns. Clin Chem 2003 (Nov);49(11):1797 – 817.
[40] Silva MF, Ruiter JP, Overmars H, et al. Complete beta-oxidation of valproate: cleavage of 3-oxovalproyl-CoA by a mitochondrial 3-oxoacyl-CoA thiolase. J Biochem 2002 (Mar 15);362(Pt 3): 755 – 60.
[41] Millington DS, Bohan TP, Roe CR, Yergey AL, Liberato DJ.Valproylcarnitine: a novel drug metabolite identified by fast atom bombardment and thermospray liquid chromatography–mass spectrometry. Clin Chim Acta 1985;145(1):69 – 76.
[42] Poplawski NK, Ranieri E, Harrison JR, Fletcher JM. Multiple acylcoenzyme A dehydrogenase deficiency: diagnosis by acyl-carnitine analysis of a 12-year-old newborn screening card. J Pediatr 1999(Jun);134(6):764 – 6.
[43] Ozand PT, Gascon GG. Organic acidurias: a review. Part 1. J Child Neurol 1991;6(3):196 – 219.
[44] Ozand PT, Gascon GG. Organic acidurias: a review. Part 2. J Child Neurol 1991;6(4):288 – 303.
[45] Chace DH, DiPerna JC, Kalas TA, Johnson RW, Naylor EW. Rapid diagnosis of methylmalonic and propionic acidemias: quantitative tandem mass spectrometric analysis of propionylcarnitine in filterpaper blood specimens obtained from newborns. Clin Chem 2001(Nov);47(11):2040 – 4.
[46] Soda H, Ohura T, Yoshida I, et al. Prenatal diagnosis and therapy for a patient with vitamin B12-responsive methylmalonic acidaemia.J Inherit Metab Dis 1995;18(3):295 – 8.
[47] Wiley V, Carpenter K, Wilcken B. Newborn screening with tandem mass spectrometry: 12 months’ experience in NSW Australia. Acta Paediatr, Suppl 1999 (Dec);88(432):48 – 51.
