Focused ion beam (FIB) milling has been adapted to thin frozen cells to enable observation of macromolecular assemblies in their native cellular context with cryogenic electron microscopy. While this approach has facilitated in situ structural studies, FIB-milling itself introduces damage that limits the amount and quality of recoverable information from cellular lamellae. The study summarized here presents an optimized low-energy milling workflow, termed Nilas, designed to minimize milling-induced damage and to produce thin biological lamellae conducive to higher-resolution in situ structural work.
Nilas is described as a low-energy milling strategy specifically tuned to reduce the physical damage caused by the ion beam during lamella preparation. The method aims to generate lamellae that retain intact macromolecular structures while reaching substantially reduced thicknesses relative to many conventional workflows. The authors emphasize that minimizing milling damage is critical to improving signal and resolution for downstream cryo-electron imaging and computational detection approaches.
Details of specific energy settings, beam currents, milling sequences, or timing parameters used in Nilas were not reported in the abstract. The report states that Nilas is compatible with common milling hardware, indicating that the approach was developed with practical accessibility in mind rather than requiring bespoke instruments.
Nilas-milled lamellae are reported to exhibit minimal FIB-milling damage. Importantly, these lamellae contain areas at or below 50 nm thickness, which is a notable target for preserving high-resolution structural information while still enabling electron transparency for imaging. The combination of reduced thickness and lowered milling damage is presented as a primary outcome of the Nilas strategy.
The abstract does not provide quantitative comparisons of damage metrics between Nilas and other milling strategies, nor does it include step-by-step protocol parameters in this summary. Such details may be present in the full preprint but were not included in the information provided here.
According to the report, Nilas-milled lamellae produce higher-resolution in situ three-dimensional reconstructions. One concrete outcome is improved recovery of ribosomal subunits from cellular sections prepared with Nilas, signifying enhanced preservation and detectability of abundant macromolecular complexes. Increased recovery of ribosomal subunits suggests improved signal for template-based and other detection methods when applied to Nilas-prepared samples.
The abstract does not enumerate the specific resolution gains (numerical resolution values) achieved in 3D reconstructions, nor does it provide the number of reconstructions or statistical measures of improvement; those data are not reported in the abstract.
A major reported advantage of Nilas is its effect on computational detection thresholds. Using two-dimensional template matching (2DTM), the authors report a reduction in the predicted minimal detectable molecular mass to approximately 220 kDa. This lowering of the detection threshold expands the range of macromolecular complexes that can be located and identified in situ.
Consistent with that predicted lowering of the detection limit, the authors report recovery of additional non-ribosomal complexes in Nilas-prepared lamellae, including RNA polymerase III, via 2DTM. This observation illustrates that Nilas can facilitate detection of less abundant or smaller complexes than previously accessible in standard FIB-milled lamellae.
The abstract does not report the full list of recovered complexes, detection rates, false-positive/false-negative rates for 2DTM, or direct comparisons to datasets generated with alternative milling methods.
The authors note that Nilas is compatible with common milling hardware. This compatibility is highlighted as an important practical advantage because it lowers the barrier for adoption across laboratories that already possess standard cryo-FIB instrumentation. The accessibility claim suggests Nilas was implemented without requiring specialized or custom-built equipment.
Specific models of FIB systems or required hardware configurations were not named in the abstract.
By extending the lower molecular-size limit detectable in situ, Nilas is presented as a step toward broader applicability of visual proteomics—the goal of locating and structurally characterizing many kinds of proteins directly within native cellular environments. Improving lamella quality and enabling detection of complexes around ~220 kDa could expand the catalog of cellular assemblies accessible to in situ structural workflows.
Because this text is drawn from the preprint abstract and metadata, several methodological and quantitative details are missing from the provided summary. The abstract does not report detailed milling parameters, quantitative damage measurements, sample numbers, statistical analyses, or full lists of recovered complexes. It also notes that the article is a preprint and has not undergone peer review. For implementation, replication, or evaluation of Nilas performance, consult the full preprint for protocol specifics, data, and validation metrics.
Competing interests were declared as none. Funders listed include the NIH Common Fund, Searle Scholars Program, Shurl and Kay Curci Foundation, Chan Zuckerberg Biohub DAF, and NSF GRFP. The full preprint and PDF were made available through bioRxiv.