This article is from
Journal of Creation 37(2):105–111, July 2023

Browse our latest digital issue Subscribe

Enantiomeric amplification of L amino acids:
part 2—chirality induced by D-sugars

by Royal Truman

We continue here with part 2 of a series of papers that claim an enantiomer excess (e.e.) of L amino acids (AAs) could arise naturally once an initial excess has somehow become available.

One approach to forming an e.e._L involves reactions mediated by chiral catalysts or chiral auxiliaries. Key experiments reported in 2017 documented the effect of chiral aldopentose sugars to mediate enantioenrichment of AA precursors in the Strecker synthesis (see pathway B in figure 1 and the results shown in table 1).¹ This differs from most of the experiments analyzed in this series, which amplify a small e.e. of an already existing AA. In this paper the pre-existing excess was present in sugars.

Adding D-pentoses to the Strecker reaction

The authors claimed they have contributed to solving the riddle of the origin of enantioselectivity of both sugars and AAs, claiming that

“This work adds to growing evidence for synergy in the etiology of the single chirality of the two most important classes of biological molecules, the sugars that make up DNA and RNA and the amino acids that form proteins.”¹

And also add:

“… this kinetic resolution may be configured so that either the amino acid resolves the sugar or, conversely, so that the sugar resolves the amino acid.”¹

Pure aminonitriles (compound III in figure 1) were used to synthesize aminoamides (IV) under very basic conditions (pH ≈ 11) in the presence of various D- or L-sugars. The e.e. of the aminoamide intermediates in the Strecker reaction typically increased for a few days under the reaction conditions before plateauing to a maximum. NMR spectroscopy indicated that several covalently bound aminonitrile-sugar species were present, responsible in some manner for the resulting e.e.¹

The results are summarized in table 1.

The enantioenrichment effect on AAs produced by D- and L-sugars was confirmed to be symmetric within measurement error (see table 2).

The e.e. depended strongly on the fraction of pure sugar present as catalyst. In this study, the authors reported an e.e. drop from 43% to 9% when the molar proportion of D-ribose:α-aminonitrile was decreased from 1:1 to 0.1:1 (see table 3).

The data from table 3 was plotted to obtain a simple polynomial curve to help visualize the enantioselectivity effect (see figure 2). The empirical equation clarifies that decreasing to more plausible values such as 0.01 relative equivalents, and using less than 100% pure D-ribose, would produce an e.e. <1%.

It will be important to keep in mind for the next section that the data in table 1 and table 2 refer to the amide intermediates before hydrolyses were performed under acidic conditions to obtain the AA (V). Without this hydrolysis step, none of the AA was obtained and the hydrolysis step IV → V did not affect the chirality.³ It must be emphasized that the data in these tables show that the biologically ‘correct’ D-ribose and D-deoxyribose produced the biologically ‘wrong’ D-AA precursors IV, whereas D-lyxose produced the desired L- precursor with high e.e._L!

This led the authors to conclude:

“Our results suggest a key role for other prebiotically common D-pentoses such as D-lyxose in mediating the emergence of L-amino acid homochirality.”¹

The importance of D-lyxose was further emphasized in this 2017 paper, with the authors stating that derivatives of this sugar had been identified on meteorite samples and had exhibited significant e.e._D.⁴

They emphasized this by pointing out that when the conversion of III → IV was mediated by a mixture of equal amounts of D-ribose and D-lyxose, or equal mixtures of the four D-pentoses, an e.e._L of IV resulted, see table 1, last two rows. We will revisit this key result in the next section when reviewing longer-term experiments reported in 2021using racemic alanine aminonitrile (III) instead of phenylalanine aminonitrile. The researchers chose not to perform the analogous experiments done in 2017 using the same mixed sugars.

Despite the contradictory effects obtained when catalyzing with different sugars, the reader is assured that

“Prebiotically plausible mixtures of natural D-sugars lead to enantioenrichment of natural L-amino acid precursors.”¹

In the next section I will present a paper published four years later, having Blackmond also as the lead author, which turns these results upside down, but I’d like to first consider the 2017 paper on its own merits.

Critique of these studies

There are several objections to the conclusions reported in 2017.¹ This paper illustrates some principles pointed out in 2022 in this journal on how bias affects which experiments to perform and how to report the results.⁵

Ideal pure starting materials were used

The chemicals used to produce sugars and AAs would have created an overwhelmingly messy mixture of materials instead of the necessary pure monomers required by DNA, RNA and proteins. Typical of these kinds of OoL experiments, the true chemical outcomes under conditions not optimized to produce the desired outcomes were not predicted. Only the desirable molecules, such as aminonitriles, were placed in clean reaction vessels in ideal proportions, protected from contaminants and interferents that lead to cross- and competing reactions.

D- and L-enantioselectivities cancelled

In OoL publications and discussions, one frequently encounters the claim that D-sugars catalyze production of L-amino acids under natural conditions. But examination of the data shows this claim is false. The data in table 2 confirmed the chemical expectation that the effects of Dand L-sugars in the enantioenrichment of aminoamides IV is symmetrical and counteracted each other for the same AA precursor.

The data in table 1 purports to show a net enhancement of L-AAs by D-sugars, especially for tryptophan. But some D-sugars increase, and others decrease the enantioenrichment of the aminoamides produced, tending to cancel out. I show in table 1 that the average effect, beginning with single pure D-sugars, ranges from e.e._L –1.2% to 4%, and we can see that the miniscule larger amount of L formed is due to only D-lyxose, a very rare sugar metabolized by some bacteria which would have been naturally present in only negligible concentration.

Deliberate experiments to emphasize catalysis using D-lyxose

Diverse sugars would differ in their effectiveness as catalysts, and so more experiments are needed than merely using an average value. It is unlikely that a considerable amount of the same D-sugar could have been present in some aqueous environment by chance; a mixture of D-sugars in various proportions would be more plausible. Which pure D-sugars should be combined for these experiments? D-ribose and D-deoxyribose (used to construct RNA and DNA) are obvious choices since the motivation of this work is to find a plausible synergistic way to naturally produce L-AAs. But table 1 shows that both these D-sugars led to AA precursors with the biologically ‘wrong’ (i.e. D) form. Performing and reporting the results of such experiments would discredit the authors’ apparent purpose of this research project. A more objective research team would have conducted these obvious experiments anyway, which would have falsified the theory of protein, RNA and DNA enantiomeric synergies.

The authors astutely included the exceptional D-lyxose in all the mixed sugar studies since it produced the desired strong L-AA enhancement in the unique case of phenylalanine. However, the most relevant sugar D-deoxyribose, the key component of DNA, was excluded from the mixture studies. Consequently, the authors were thereby able to report a net selectivity of L-AA precursors in their results. Those with a more biologically relevant objective would surely have also reported the results from experiments using mixtures of the two most relevant D-sugars instead of focusing on the biologically irrelevant D-lyxose.

The authors justify using D-lyxose in these studies by stating:

“… enantioenriched derivatives of the biologically rare lyxose were found in similar abundance to derivatives of ribose on two separate meteorites.”¹

Note the word derivatives in the above quote. Lyxose itself was not delivered to Earth but had to be chemically extracted in a laboratory under vigorous hydrolysis conditions. They also did not mention that Cooper and Rios, who authored the paper they quoted, reported finding a possible trace amount of D-lyxonic acid in the soil in the immediate area where the Murchison sample was found, implying this e.e. may have originated from terrestrial bacteria.⁴ This is quite reasonable, considering that most chiral meteoritic compounds have been found to be racemic mixtures.⁶

But what Cooper and Rios reported after chemical processing of material dissolved out of the two meteorite samples was D-lyxonic acid. No justification was provided for using D-lyxose instead in the experiments as the sugar catalyst, nor was this substitution even mentioned.

It must be emphasized again that the sugars found in meteorites are bound in large complex substances which only existed as sugars after being freed using vigorous chemical liberation conditions. This makes it unrealistic to accumulate concentrated sugars from meteorites under natural conditions. One must avoid the error of thinking that perhaps exposure to higher temperatures and millions of years could have eventually freed the bound compounds. Under such circumstances the AAs and sugars would have been lost e.e. through racemization.^6-9

Multiple synthetic steps under incompatible conditions

In the reaction overview shown in figure 1, the intended products were AAs. Conversion of aminonitrile III → IV was carried out at a pH ≈ 11, but hydrolysis of IV to form V did not occur under the reaction conditions for the phenylalanine pathway:

“… further hydrolysis to the amino acid was not observed and was found to require more forcing conditions.”²

Amides can be hydrolyzed in either acidic or basic conditions. In laboratories and chemical manufacturing, alkaline hydrolysis involves heating the amide with hot aqueous sodium or potassium hydroxide, but these conditions would rapidly racemize any biologically relevant AA. Therefore, acidic conditions would make more sense, using sulfuric or hydrochloric acid, and temperatures of about 100°C for several hours.¹⁰

Unlike these highly skilled chemists, nature would not foresee that the chiral centre must not be damaged by removing the proton at the α-carbon under the highly alkaline conditions of the preceding reaction step. Over hundreds or thousands of years the AAs produced would have racemized V. Alternatively, it is unreasonable to demand chance to transfer the aminoamide IV from an environment with pH near 11 to a very acidic one at just the right time.

To reiterate, the reaction steps shown in figure 1 were not one-pot reactions. Each intermediate was synthesized under a separate set of ideal conditions and reagents. Typical of OoL experiments, the intermediates were purified, then isolated, and then new laboratory setups used with different reagents and conditions for the next step. Here, as in almost all such studies, pure intermediates (like aminonitriles III) were simply purchased, since organic chemists know that beginning with aldehydes like I mixed together in water with all the other reagents needed such as NH₃, HCl, NaOH, KCN and sugars would produce amino acids V in at best trace amounts mixed with a complex mixture of other chemicals.

Therefore, the researchers introduced pure D-sugars precisely at the beginning of the key reaction of aminonitriles III shown in figure 1, thus avoiding the complex and polyfunctional mixtures resulting from exposure of these sugars to the other key reactants I, II, and V.

Unrealistic concentrations and stoichiometries are used

For the key step shown in figure 1, 0.25 M of an aminonitrile III was mixed with 0.50 M D-sugar in water. Aminoamide products IV were produced in 9–29% isolated yield.¹ This was accomplished with the help of 0.25 M NaOH at 22–24 °C. Ala-III and Phe-III were reacted for 7 days and Trp-III for 5 days.¹

However, these are massively unrealistically high concentrations of III plus D-sugar under natural conditions, which would have had to be present concurrently during an appropriate time interval. It is legitimate to use optimized concentrations and ratios for research convenience, but one should then extrapolate to natural conditions. Even assuming the most extremely optimistic coincidence of conditions, nature does not provide hermetical constraining volumes with just the right reagents in ideal proportions along with a continuous mixing mechanism (as with the laboratory’s magnetic stirrer).

To illustrate, one could place a few drops of concentrated III and D-sugar in close proximity in a container the size of an Olympic swimming pool, and then determine a week later how much IV is found throughout the entire volume. Given the statistical impossibility of finding twice as much D-sugar as III in a nearby volume, and the relative ease of hydrolysis of III under fluctuating temperatures and high concentration of bases, no detectable e.e._L of product IV would be likely to form naturally.

Long times would be necessary

A pH ≈ 11 (required to produce aminonitriles IV) would racemize the target L-amino acids V. Perhaps a lower pH could be used to form IV instead. The authors tested a stoichiometry of 2:1 D-ribose:III with deionized water (pH = 7) and also 0.0001 M NaOH (effective pH = 10), obtaining e.e._L values of only 35% and 36% respectively (see table 4 in ref. 1). This is half the e.e._L of 70% we saw in table 2 even though the researchers had increased the reaction time from 7 to 35 days. Greater separation through diffusion would result (recall the example of using a swimming pool above), resulting in reactant concentrations below threshold levels within minutes. This would have led to no catalytic effect from the D-sugar.

We could continue to extrapolate into more plausible conditions by also evaluating the effect of lowering the proportion of D-sugar:III, based on the results from table 3 and figure 2. Recall, however, that these are trapped in a close container and mixed with a magnetic stirrer, rapidly forcing the reagents into close contact. The rational conclusion is that the e.e._L reported is only an artifact of the carefully planned experiments. Any L-AA V formed from IV would do so exceedingly slowly and would racemize in water. Not only would the theoretical e.e. generated be negligible at any location, but it would also be present in very low concentration at any time.

Some systematic experiments monitoring aminoamide e.e. values for longer than merely a few days would have been helpful to extrapolate to prebiotic natural conditions and to demonstrate the impossibility of the proposed mechanisms. One would expect the aminoamides to racemize, especially in the presence of a strong base. From table 4 of the paper, we observe that the e.e._D for Phe-precursor decreased from 46% to 36% when, instead of pure water as solvent, 0.00010 M NaOH was used at 37°C after 35 days.¹ Thus, presumably an initial e.e. produced in some manner would racemize over time (weeks), especially after thousands or millions of years.

Even though the aminonitrile III disappeared in under 4 hours, the e.e. of Phe-IV increased from a racemic mixture at the outset of the reaction and continued to rise for nearly 1 week, (see figure 3).

Enantiomeric excess for alanine in the presence of D-ribose

The focus of the 2017 paper was on the synthetic pathway for phenylalanine.¹ In a later paper published in 2021, the alanine pathway was examined with results inimical to the conclusions of the earlier paper.² We see that now to obtain an e.e._L for alanine, the presence of D-lyxose is the last thing the OoL researchers would want, since it catalyzed formation of D-alanine instead!

The Ala aminonitrile III (instead of Phe, as before) now produced L-V in the presence of D-ribose, see figure 4.² The difference was that Ala-IV hydrolyzed to Ala-V (whereas Phe-IV had not) and the sugar enhancer changed the chirality during hydrolysis. Recall that in the 2017 report, hydrolysis to form Phe-IV required a separate step, using strongly acidic conditions which did not change the chirality.

Figure 5 shows the development of e.e._D for Ala-IV and Ala-V over time. Ala-IV was nearly racemic initially but eventually formed a 13% e.e._D, whereas Ala-V, initially c. 50% e.e._L, rose to 68% e.e._L.

To recap, table 1 showed, for the 2017 study, that racemic Phe-III in the presence of D-ribose resulted primarily in D-Phe but in primarily the biological L-Phe in the presence of D-lyxose.

The 2021 study produced the opposite results. Racemic Ala-III in the presence of D-ribose resulted primarily in L-Ala and primarily D-Ala in the presence of D-lyxose. Embarrassingly, D-lyxose was revealed to be detrimental to producing L-AAs, contradicting the break-through discoveries in the 2017 paper.

The chemically simpler Ala or its derivatives would be much easier to form naturally than Phe and is/are found in much higher proportion in proteins. Since rac-Ala-III in the presence of D-lyxose would inevitably lead to D-Ala (the ‘wrong’ AA) without any change in reaction conditions, an OoL researcher might wish to switch strategy and now downplay the existence of this sugar. Unfortunately, this would then eliminate the most effective sugar known to obtain L-Phe from racemic Phe-III.

In the 2017 paper, mixed sugars which always contained D-lyxose with Ph-III were heralded enthusiastically as a means for obtaining L-Phe, but the analogous experiments using the same mixtures with Ala-III were not carried out later, perhaps once the anticipated results became known. This is an example of how the choice of research to be conducted predetermined what readers learned, and misled them into thinking that the bulk of the evidence supported evolution.⁵

Perhaps the most important fact about these experiments is that the e.e. found is only a temporary effect. The kinetic and NMR analysis demonstrated that when an excess of L enantiomer was obtained, much of the potential D-AA remained sequestered temporarily and would be automatically produced later.

Figure 6 shows how the proportion of D-Ribose and NaOH strongly affected how much L-Ala-V formed. The green points represent excess of L, and the right axis represents –e.e._D values. One notes how, within mere days, the initial e.e._L began to decrease and would soon lead to a racemic Ala mixture.

Clearly the e.e. effects reported were merely a laboratory artifact obtained under unnatural conditions, caused primarily by exposure to extremely high proportions of the D-sugar present.

The data shown in figure 6 reveals that a researcher could have obtained high or low e.e. values and increasing or decreasing trends with time depending on the exact conditions and duration used. Unusual non-equilibrium short-term effects can be obtained by selecting suitable parameters. But naturalist models require deep time. The relevant consideration is what would result after thousands or millions of years with no deliberate guidance. Under the already implausible conditions selected for the data found in figure 6, waiting a few months would have revealed no e.e. present at all!

Additional critical observations

Where would the high concentration of a sugar such as D-ribose have come from? The scenario favoured by the OoL community involves polymerization of formaldehyde (the formose reaction), against which chemistry professor Shapiro has levelled devastating chemical objections.¹¹ In the words of Weiss et al.:¹²

“The formose product can be regarded as a carbohydrate analog of petroleum, in that it contains so many carbohydrates of varying molecular weight and isomeric structure.”

Formaldehyde has little tendency to react with itself to form carbon–carbon bonds and instead much of the formaldehyde is used by other chemical processes such as the Cannizzaro reaction.¹³ In fact, formaldehyde reacts much more easily with many other substances used in various prebiotic proposals, including ammonia, amines, amides, imides, aminonitriles, urethanes, and urea.¹²

Reid and Orgel were not able to produce any sugar-like substances from formaldehyde unless the concentration was at least 0.5 M, a concentration far, far higher than could arise naturally.¹⁴ Miller and Orgel have also stated that it would be problematic for a high enough concentration of formaldehyde to have existed on a primitive Earth to produce even trace amounts of D-ribose.¹⁵ Once some sugar has formed via the formose reaction, further reactions would have to be quickly stopped (e.g. within about 10–13 min) to prevent decomposition.^14,16

OoL researchers Reid and Orgel therefore have concluded:¹⁴

“We do not believe that the formose reaction as we and others have carried it out is a plausible model for the prebiotic accumulation of sugars.”

And Shapiro adds:¹²

“Little has happened in the decades since that report to alter the above judgement.”

Another issue is that D-ribose isomerizes easily. Tewari and Goldberg reported that an initial sample of D-ribose formed a mixture of 75% D-arabinose, 6% D-ribulose, and that only 19% D-ribose remained within a few weeks at 25°C and pH 7.¹⁷

Final comments

The experiments discussed above were not conducted under a consistent set of conditions to form D-ribose, D-deoxyribose, and the 19 L biological amino acids. Some D-sugars led to an e.e. of L-AAs, whereas other sugars enhanced the opposite, and the effects were not inconsistent for the various AAs. Without the guidance of the researchers, over time racemic mixtures of AAs would have formed.