RPLC-HRMS Reveals PLGA Block-Length Differences

In a recently published paper in the journal Macromolecules,¹ Masashi Serizawa and Andrea Gargano, researchers at the Van ‘t Hoff Institute for Molecular Sciences (HIMS, University of Amsterdam), present a “polymer sequencing” approach to analyze the block-length composition of poly(lactic-co-glycolic) acid (PLGA) copolymers used for drug release. The method consists of controlled chemical degradation followed by reversed-phase liquid chromatography and high-resolution mass spectrometry (RPLC-HRMS). The generated PLGA oligomers were used to infer the block-length distribution in the original polymer. This approach was used to reveal differences in block length between PLGA copolymers with identical chemical composition but exhibiting different solubility in acetone. LCGC International spoke to Serizawa and Gargano about their work.

What problem does your work solve in PLGA characterization?

PLGA is widely used as a polymer for drug delivery in nanoparticle formulations, yet current analytical methods provide only limited insight into its sequence heterogeneity and block-length distribution (BLD). Even polymers with identical composition and molecular weight can show very different performance due to differences in blockiness.

In this work, we introduce a polymer sequencing strategy to characterize PLGA microstructure and obtain information on the dimer alternating structures present in the co-polymer and its BLD (Figure 1).

The analytical approach developed was then applied to demonstrate that differences in block-length distributions led to distinct properties, as exemplified by the study of polymer solubility, even for polymers with identical average chemical compositions. In particular, more blocky structures have lower solubility in acetone (due to GA-rich crystalline blocks), whereas more random structures have higher solubility.

Acetone solubility is considered a structurally sensitive and industrially relevant probe, as differences directly reflect variations in PLGA sequence blockiness, including degradation and release profiles. State-of-the-art analytical methods, such as NMR, often provide only limited insight into the various properties of PLGA polymers.

What is the key idea behind your analytical approach?

Our method combines: (i) controlled chemical degradation, (ii) RPLC–HRMS analysis and (iii) computational reconstruction of the block-length distribution of the original polymer (SWAMP-MS).² The controlled degradation converts high-MW PLGA into oligomers (< ~1 kDa) suitable for LC-MS primarily via hydrolysis of glycolic acid residues. These oligomers are then used to reconstruct the block-length distribution of the original polymer. We use two complementary degradation conditions: a shorter degradation (30 min) that releases longer oligomers to study residual dimer-alternating motifs and a longer degradation (3 hr) to investigate the block-length distribution (Figure 2). This dual approach allows us to access different levels of sequence information.

The authors used controlled chemical degradation before LC-MS analysis. Why is this sample preparation step necessary for high-molecular-weight PLGA polymers?

Direct analysis of polymers above 10 kDa by LC–MS is not feasible due to their high molecular weight, structural complexity, and limited detectability. In practice, LC-ESI-MS of polydisperse polymers such as PLGA yields a convoluted mass spectrum that cannot be used to extract molecular information about the polymer.

Therefore, we used controlled degradation of the polymer as a sample preparation step to obtain oligomers from polymers with molecular weights over 100 kDa. The chemically degraded oligomers are lower-molecular-weight fragments (< 2 kDa) that are more amenable to LC–MS analysis. These fragments retain essential information about the original microstructure, and the polymer sequence can subsequently be reconstructed using a dedicated algorithm.²

In this work, we found that the degree of degradation critically affects the type of structural information obtained. Therefore, we employed two different degradation conditions with distinct extents of chain cleavage. This dual approach enabled us to capture complementary structural features, specifically both block structures and dimer-alternating-sequence motifs.

In summary, for complex polymers such as PLGA, controlled chemical degradation is a necessary sample preparation step that simplifies the system while preserving key structural information. When combined with computational reconstruction, this approach provides an effective means of extracting sequence-level microstructural details that are otherwise inaccessible by direct LC–MS analysis of the intact polymer.

Why was reversed-phase liquid chromatography (RPLC) selected for the separation of PLGA degradation oligomers prior to HRMS analysis?

Reversed-phase liquid chromatography (RPLC) was selected because of its high chemical selectivity for separating species based on structural differences, including chain length, lactide/glycolide composition, and end groups. As shown in Figure 3, oligomers with up to 12 monomers obtained by chemical degradation are resolved and elute at different times depending on whether their end groups are ester or acidic. Finally, the RPLC method used (employing water and ACN with 0.1% FA) is well compatible with HRMS.

With your method, was it possible to separate PLGA oligomers that differ only in monomer sequence but have identical molecular weights?

We did not investigate the separation of PLGA isomers that differ in the organization of the LA-GA units (e.g., in a trimer LA-GA-LA vs. LA-LA-GA), although this information may provide further insight into the blockiness of the polymer. Such separations are challenging as these oligomers often exhibit very similar physicochemical properties. The relatively steep gradient we adopted in our analysis (8% B/min) on a relatively short column (50 mm) did not allow this. Shallower gradient on a longer column and/or the use of advanced MS methods, such as ion mobility, may provide access to this information and be a potential source of further insights.

How could extracted ion chromatograms (EICs) be used to investigate the abundance of specific PLGA oligomer sequences and block-length distributions?

In total ion chromatograms (TICs), all oligomers are collectively detected after ionization, providing an overall profile of the sample. However, for structural analysis purposes, it is often more informative to focus on specific subsets of oligomers, such as PLA-rich or PGA-rich oligomer series, rather than considering the entire distribution.

EICs were used in our work to get insights into sequence characteristics, such as dimer-alternating arrangements and differences in blockiness. In our work we used this strategy to gain insight into the presence of LA and GA dimers in the polymer sequence. When dimers were present, the intensity of oligomers of the same chemical series (e.g., oligomers with one GA unit and multiple LA, Figure 4) with even repeating units was significantly lower than the even oligomers, giving a characteristic “sawtooth” intensity pattern, whereas randomized polymers gave a more regular intensity distribution.

The study compares acetone-soluble and acetone-insoluble PLGA fractions. How could LC-HRMS data be used to correlate chromatographic profiles with differences in polymer blockiness and solubility?

The correlation between polymer blockiness and solubility is strongly influenced by the presence of crystalline glycolide (GA) blocks; in general, longer GA block lengths tend to reduce solubility, leading to the formation of acetone-insoluble fractions.

Ideally, LC–HRMS would allow direct observation of GA block lengths, enabling a straightforward correlation between chromatographic profiles, blockiness, and solubility. While intact GA sequences are destroyed in our method during controlled hydrolysis, they can be indirectly quantified by measuring the surviving LA blocks. Due to mass conservation constraints, the detection of long LA blocks inherently confirms that long GA domains existed elsewhere in the copolymer chain, providing a clear structural indicator of blocky sequence heterogeneity over a randomized distribution (Figure 5).

This principle was employed in the SWAMP algorithm to reconstruct the GA block structures that were not fully detectable in the LC–HRMS data. This approach enables indirect but effective evaluation of block-length distributions and their relationship to solubility.

What potential sources of analytical bias or error could arise during the chemical degradation, chromatographic separation, and mass spectrometric detection of PLGA oligomers?

As with any multi-step analytical workflow, potential biases can arise during sample preparation, chromatographic separation, and mass spectrometric detection. However, the key question is whether these factors affect the ability to reconstruct PLGA microstructure.

The most important consideration is the chemical degradation step, since the structural information extracted by our method is derived from the resulting oligomers. We therefore carefully controlled the degradation conditions to ensure reproducible oligomer generation and meaningful comparison between samples.

Chromatographic separation and HRMS detection may also influence the measured abundance of individual oligomers through factors such as co-elution or differences in ionization efficiency. However, the combination of high-resolution chromatographic separation, accurate-mass identification, and analysis of large oligomer populations helps minimize the impact of these effects on the overall interpretation.

Importantly, our conclusions are not based on a single oligomer species but on trends observed across many degradation products. This makes the method less sensitive to variability associated with individual chromatographic or mass spectrometric signals.

Overall, while analytical limitations should always be considered, the workflow provides a robust and reproducible means of comparing PLGA microstructures and reconstructing block-length distributions.

If two PLGA samples exhibit the same NMR-derived randomness index but different block-length distributions, how could LC-HRMS chromatographic and mass spectral data reveal these structural differences?

Even when two PLGA samples exhibit the same NMR-derived randomness index, differences in block-length distributions can lead to distinct degradation behaviors, as reported in previous studies. Variations in block length influence hydrolysis kinetics, resulting in different oligomer populations after controlled chemical degradation.

The NMR-derived randomness index is an ensemble-averaged parameter that reflects the overall frequency of LA–LA, LA–GA, and GA–GA linkages. As a result, PLGA samples with substantially different block-length distributions may nevertheless exhibit similar randomness indices.

As a result, LC–HRMS analysis of the degraded samples produces distinct mass spectral profiles, reflecting these differences in sequence distribution. In addition to mass spectral information, chromatographic behavior can provide insight, as oligomers derived from more blocky structures tend to exhibit different retention-time distributions than those from more uniformly distributed sequences.

To visualize these differences more clearly, oligomer distributions obtained by LC–HRMS can be represented as contour plots (e.g., number of LA units vs. number of GA units). Such plots enable direct comparison of compositional distributions between samples, highlighting differences in sequence heterogeneity and block-length distributions that are not detectable by NMR alone.

Therefore, while NMR provides valuable information on the average sequence structure of PLGA, LC–HRMS offers a more detailed view of the underlying sequence distribution. The two techniques are complementary, but LC–HRMS can reveal differences in block-length distribution that may remain hidden when samples are characterized solely by NMR-derived randomness indices.

Thus, LC–HRMS, combined with appropriate data visualization, allows discrimination between polymer samples with identical average randomness but different underlying sequence distributions.

References

1. M. Serizawa; D. Snijders; S. van Berkel, et al. Quantifying Poly(lactide-co-glycolide) Residual Dimer-Alternating Structure and Block-Length Distribution Using Controlled Chemical Degradation and Reverse-Phase Liquid Chromatography–Mass Spectrometry Analysis. Macromolecules 2026, 59 (7), 4473–4484. DOI: 10.1021/acs.macromol.5c03061
2. Bos, T. S.; van den Hurk, R. S.; Mengerink, Y. et al. Determination of Copolymer Block-Length Distributions Using Fragmentation Data Obtained from Tandem Mass Spectrometry. Macromolecules 2025, 58 (13), 6430–6439. DOI: 10.1021/acs.macromol.5c00297